Datasets:
id string | dataset string | full_name string | version string | language string | programming_language string | task_type string | difficulty string | title string | domain string | instruction string | code_context string | intent string | constraints list | reference_solution string | explanation string | test_cases list | quality_signals dict | evaluation dict | curation dict | tags list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
J6-CFI-HQ-20K-00001 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | rust | refactoring | medium | Refactoring — Tokio Task Coordinator (rust) | payment_systems | Refactor the rust Tokio task coordinator in a payment systems module. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: 6b6367a73d. | // case: 6b6367a73d
fn normalize_6b6367a73d(items: Vec<Option<String>>, strict: bool) -> Result<Vec<String>, String> {
let mut result = Vec::new();
for raw in items {
match raw {
Some(value) => {
let cleaned = value.trim().to_lowercase();
if !cleaned.is_empty(... | Produce a high-quality refactoring response for a rust Tokio task coordinator while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case 6b6367a73d, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a rust Tokio task coordinator, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_6b6367a73d",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_6b6367a73d",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "6b6367a73d"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete rust Tokio task coordinator.",
"Instruction contains explicit behavioral constraints."... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"rust",
"payment_systems",
"medium"
] |
J6-CFI-HQ-20K-00002 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | debugging | hard | Debugging — Union Error Model (typescript) | security | Debug the typescript union error model inside a production-facing developer tool. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: b676166168. | // case: b676166168
type NormalizeOptions_b676166168 = { strict?: boolean };
export function normalize_b676166168(items: unknown[], options: NormalizeOptions_b676166168 = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality debugging response for a typescript union error model while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case b676166168, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a typescript union error model, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_b676166168",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_b676166168",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "b676166168"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript union error model.",
"Instruction contains explicit behavioral constraints.... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"typescript",
"security",
"hard"
] |
J6-CFI-HQ-20K-00003 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | php | code_explanation | easy | Code Explanation — Pagination Renderer (php) | developer_tools | Explain the php pagination renderer while preserving an existing workflow. Describe the data flow, one realistic failure mode, and what should be tested next. Case: d90ac6db03. | <?php
// case: d90ac6db03
function normalize_d90ac6db03(array $items, bool $strict = false): array {
$result = [];
foreach ($items as $raw) {
if ($raw === null) {
if ($strict) {
throw new InvalidArgumentException('Null values are not allowed');
}
conti... | Produce a high-quality code explanation response for a php pagination renderer while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Explain what the code does before suggesting changes",
"Call out one realistic failure mode",
"Avoid rewriting the code",
"Keep the explanation concise and technical"
] | For case d90ac6db03, Walk through input handling, transformation, and output construction without rewriting the code. Mention one realistic failure mode and a focused test. | This high-quality synthetic sample evaluates whether a coding assistant can handle code explanation for a php pagination renderer, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_d90ac6db03",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_d90ac6db03",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "d90ac6db03"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_explanation",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete php pagination renderer.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_explanation",
"php",
"developer_tools",
"easy"
] |
J6-CFI-HQ-20K-00004 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | php | code_generation | easy | Code Generation — Pagination Renderer (php) | testing | Implement a php pagination renderer for a realistic service boundary. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 83788d2126. | <?php
// case: 83788d2126
function normalize_83788d2126(array $items, bool $strict = false): array {
$result = [];
foreach ($items as $raw) {
if ($raw === null) {
if ($strict) {
throw new InvalidArgumentException('Null values are not allowed');
}
conti... | Produce a high-quality code generation response for a php pagination renderer while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 83788d2126, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a php pagination renderer, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_83788d2126",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_83788d2126",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "83788d2126"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete php pagination renderer.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"php",
"testing",
"easy"
] |
J6-CFI-HQ-20K-00005 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | code_generation | medium | Code Generation — Stream Helper (java) | payment_systems | Implement a java stream helper without changing unrelated behavior. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 9a4318b062. | // case: 9a4318b062
public List<String> normalize_9a4318b062(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality code generation response for a java stream helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 9a4318b062, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a java stream helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_9a4318b062",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_9a4318b062",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "9a4318b062"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java stream helper.",
"Instruction contains explicit behavioral constraints.",
"Re... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"java",
"payment_systems",
"medium"
] |
J6-CFI-HQ-20K-00006 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | go | sql_data_tasks | hard | Sql Data Tasks — Cli Command (go) | web_frontend | Improve the SQL CLI command in a web frontend module. Preserve result semantics, handle NULL values, and discuss index or scan implications. Case: 9fa196a4e6. | // case: 9fa196a4e6
func Normalize_9fa196a4e6(items []string, strict bool) ([]string, error) {
result := make([]string, 0, len(items))
for _, raw := range items {
value := strings.TrimSpace(raw)
if value == "" {
if strict { return nil, fmt.Errorf("empty values are not allowed") }
... | Produce a high-quality sql data tasks response for a go CLI command while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Preserve result semantics",
"Handle NULL values deliberately",
"Use explicit joins or filters",
"Discuss index or scan implications"
] | For case 9fa196a4e6, Make joins and filters explicit, define NULL handling, preserve aggregation semantics, and describe the index or query shape that reduces scanning. | This high-quality synthetic sample evaluates whether a coding assistant can handle sql data tasks for a go CLI command, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_9fa196a4e6",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_9fa196a4e6",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "9fa196a4e6"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_join",
"missin... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete go CLI command.",
"Instruction contains explicit behavioral constraints.",
"Refere... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"sql_data_tasks",
"go",
"web_frontend",
"hard"
] |
J6-CFI-HQ-20K-00007 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | algorithmic_reasoning | medium | Algorithmic Reasoning — Zod Request Validator (typescript) | performance | Solve the typescript Zod request validator problem inside a production-facing developer tool. Explain the invariant, time complexity, space complexity, and boundary behavior. Case: 5b937bb59d. | // case: 5b937bb59d
type NormalizeOptions_5b937bb59d = { strict?: boolean };
export function normalize_5b937bb59d(items: unknown[], options: NormalizeOptions_5b937bb59d = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality algorithmic reasoning response for a typescript Zod request validator while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"State the algorithmic invariant",
"Give time and space complexity",
"Handle empty and boundary inputs",
"Avoid brute force when a better approach is expected"
] | For case 5b937bb59d, Describe the invariant, choose the simplest algorithm that meets the complexity target, prove boundary behavior, and state time and space complexity. | This high-quality synthetic sample evaluates whether a coding assistant can handle algorithmic reasoning for a typescript Zod request validator, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_5b937bb59d",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_5b937bb59d",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "5b937bb59d"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_complexity",
"... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript Zod request validator.",
"Instruction contains explicit behavioral constrai... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"algorithmic_reasoning",
"typescript",
"performance",
"medium"
] |
J6-CFI-HQ-20K-00008 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | debugging | hard | Debugging — Repository Query Adapter (java) | automation | Debug the java repository query adapter while preserving an existing workflow. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: 00c1bc7d91. | // case: 00c1bc7d91
public List<String> normalize_00c1bc7d91(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality debugging response for a java repository query adapter while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case 00c1bc7d91, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a java repository query adapter, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_00c1bc7d91",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_00c1bc7d91",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "00c1bc7d91"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java repository query adapter.",
"Instruction contains explicit behavioral constraints... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"java",
"automation",
"hard"
] |
J6-CFI-HQ-20K-00009 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | html_css | code_generation | hard | Code Generation — Empty State Component (html_css) | payment_systems | Implement a html_css empty state component for a realistic service boundary. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: a2241e1571. | /* case: a2241e1571 */
.component-a2241e1571 {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(18rem, 1fr));
gap: 1rem;
}
.component-a2241e1571 .card {
min-width: 0;
padding: 1rem;
} | Produce a high-quality code generation response for a html_css empty state component while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case a2241e1571, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a html_css empty state component, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_a2241e1571",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_a2241e1571",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "a2241e1571"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete html_css empty state component.",
"Instruction contains explicit behavioral constraint... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"html_css",
"payment_systems",
"hard"
] |
J6-CFI-HQ-20K-00010 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | code_generation | easy | Code Generation — Typed Api Client (typescript) | observability | Implement a typescript typed API client without changing unrelated behavior. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: d1b3bf1a17. | // case: d1b3bf1a17
type NormalizeOptions_d1b3bf1a17 = { strict?: boolean };
export function normalize_d1b3bf1a17(items: unknown[], options: NormalizeOptions_d1b3bf1a17 = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality code generation response for a typescript typed API client while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case d1b3bf1a17, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a typescript typed API client, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_d1b3bf1a17",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_d1b3bf1a17",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "d1b3bf1a17"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript typed API client.",
"Instruction contains explicit behavioral constraints."... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"typescript",
"observability",
"easy"
] |
J6-CFI-HQ-20K-00011 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | cpp | code_explanation | easy | Code Explanation — String Token Parser (cpp) | performance | Explain the cpp string token parser in a performance module. Describe the data flow, one realistic failure mode, and what should be tested next. Case: 5bb3ca419a. | // case: 5bb3ca419a
vector<string> normalize_5bb3ca419a(const vector<string>& items) {
vector<string> result;
for (const auto& raw : items) {
string value = trim(raw);
if (!value.empty()) result.push_back(to_lower(value));
}
return result;
} | Produce a high-quality code explanation response for a cpp string token parser while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Explain what the code does before suggesting changes",
"Call out one realistic failure mode",
"Avoid rewriting the code",
"Keep the explanation concise and technical"
] | For case 5bb3ca419a, Walk through input handling, transformation, and output construction without rewriting the code. Mention one realistic failure mode and a focused test. | This high-quality synthetic sample evaluates whether a coding assistant can handle code explanation for a cpp string token parser, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_5bb3ca419a",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_5bb3ca419a",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "5bb3ca419a"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_explanation",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete cpp string token parser.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_explanation",
"cpp",
"performance",
"easy"
] |
J6-CFI-HQ-20K-00012 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | algorithmic_reasoning | expert | Algorithmic Reasoning — Cache Service (java) | cli_tools | Solve the java cache service problem inside a production-facing developer tool. Explain the invariant, time complexity, space complexity, and boundary behavior. Case: c68e19fc0a. | // case: c68e19fc0a
public List<String> normalize_c68e19fc0a(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality algorithmic reasoning response for a java cache service while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"State the algorithmic invariant",
"Give time and space complexity",
"Handle empty and boundary inputs",
"Avoid brute force when a better approach is expected"
] | For case c68e19fc0a, Describe the invariant, choose the simplest algorithm that meets the complexity target, prove boundary behavior, and state time and space complexity. | This high-quality synthetic sample evaluates whether a coding assistant can handle algorithmic reasoning for a java cache service, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_c68e19fc0a",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_c68e19fc0a",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "c68e19fc0a"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_complexity",
"... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java cache service.",
"Instruction contains explicit behavioral constraints.",
"Re... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"algorithmic_reasoning",
"java",
"cli_tools",
"expert"
] |
J6-CFI-HQ-20K-00013 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | go | code_generation | easy | Code Generation — Repository Method (go) | database | Implement a go repository method while preserving an existing workflow. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 43747cdffc. | // case: 43747cdffc
func Normalize_43747cdffc(items []string, strict bool) ([]string, error) {
result := make([]string, 0, len(items))
for _, raw := range items {
value := strings.TrimSpace(raw)
if value == "" {
if strict { return nil, fmt.Errorf("empty values are not allowed") }
... | Produce a high-quality code generation response for a go repository method while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 43747cdffc, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a go repository method, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_43747cdffc",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_43747cdffc",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "43747cdffc"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete go repository method.",
"Instruction contains explicit behavioral constraints.",
"... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"go",
"database",
"easy"
] |
J6-CFI-HQ-20K-00014 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | sql | refactoring | hard | Refactoring — Revenue Aggregation Query (sql) | performance | Refactor the sql revenue aggregation query for a realistic service boundary. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: 1860b0f2f3. | -- case: 1860b0f2f3
SELECT account_id,
COUNT(*) AS total_events,
MAX(created_at) AS last_event_at
FROM events_1860b0
WHERE created_at >= CURRENT_DATE - INTERVAL '30 days'
AND account_id IS NOT NULL
GROUP BY account_id
ORDER BY total_events DESC; | Produce a high-quality refactoring response for a sql revenue aggregation query while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case 1860b0f2f3, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a sql revenue aggregation query, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_1860b0f2f3",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_1860b0f2f3",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "1860b0f2f3"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete sql revenue aggregation query.",
"Instruction contains explicit behavioral constraints... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"sql",
"performance",
"hard"
] |
J6-CFI-HQ-20K-00015 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | test_generation | hard | Test Generation — Exception Handler (java) | database | Generate tests for the java exception handler without changing unrelated behavior. Cover success, boundary, failure, and regression cases with clear assertions. Case: af7c7f66c7. | // case: af7c7f66c7
public List<String> normalize_af7c7f66c7(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality test generation response for a java exception handler while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Cover success, boundary, and failure cases",
"Use behavior-focused test names",
"Avoid tests that depend on implementation details",
"Include at least one regression-style case"
] | For case af7c7f66c7, Build tests for normal input, boundary input, malformed input, and a regression scenario. Assert externally visible behavior rather than private implementation details. | This high-quality synthetic sample evaluates whether a coding assistant can handle test generation for a java exception handler, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_af7c7f66c7",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_af7c7f66c7",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "af7c7f66c7"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"weak_assertions",
"m... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java exception handler.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"test_generation",
"java",
"database",
"hard"
] |
J6-CFI-HQ-20K-00016 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | test_generation | easy | Test Generation — Batch Import Service (java) | admin_panels | Generate tests for the java batch import service in a admin panels module. Cover success, boundary, failure, and regression cases with clear assertions. Case: dd53d0f34f. | // case: dd53d0f34f
public List<String> normalize_dd53d0f34f(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality test generation response for a java batch import service while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Cover success, boundary, and failure cases",
"Use behavior-focused test names",
"Avoid tests that depend on implementation details",
"Include at least one regression-style case"
] | For case dd53d0f34f, Build tests for normal input, boundary input, malformed input, and a regression scenario. Assert externally visible behavior rather than private implementation details. | This high-quality synthetic sample evaluates whether a coding assistant can handle test generation for a java batch import service, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_dd53d0f34f",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_dd53d0f34f",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "dd53d0f34f"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"weak_assertions",
"m... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java batch import service.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"test_generation",
"java",
"admin_panels",
"easy"
] |
J6-CFI-HQ-20K-00017 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | code_review | hard | Code Review — Rest Validation Controller (java) | payment_systems | Review the java REST validation controller inside a production-facing developer tool. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: 3e1edba572. | // case: 3e1edba572
public List<String> normalize_3e1edba572(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality code review response for a java REST validation controller while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case 3e1edba572, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a java REST validation controller, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_3e1edba572",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_3e1edba572",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "3e1edba572"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java REST validation controller.",
"Instruction contains explicit behavioral constrain... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"java",
"payment_systems",
"hard"
] |
J6-CFI-HQ-20K-00018 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | refactoring | hard | Refactoring — Dictionary Normalizer (python) | observability | Refactor the python dictionary normalizer while preserving an existing workflow. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: 951152a836. | # case: 951152a836
def normalize_951152a836(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality refactoring response for a python dictionary normalizer while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case 951152a836, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a python dictionary normalizer, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_951152a836",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_951152a836",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "951152a836"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python dictionary normalizer.",
"Instruction contains explicit behavioral constraints.... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"python",
"observability",
"hard"
] |
J6-CFI-HQ-20K-00019 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | csharp | refactoring | medium | Refactoring — Background Worker (csharp) | payment_systems | Refactor the csharp background worker for a realistic service boundary. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: 5cca8ef21d. | // case: 5cca8ef21d
public IEnumerable<string> Normalize_5cca8ef21d(IEnumerable<object?> items, bool strict = false)
{
foreach (var raw in items)
{
if (raw is null)
{
if (strict) throw new ArgumentException("Null values are not allowed");
continue;
}
var v... | Produce a high-quality refactoring response for a csharp background worker while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case 5cca8ef21d, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a csharp background worker, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_5cca8ef21d",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_5cca8ef21d",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "5cca8ef21d"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete csharp background worker.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"csharp",
"payment_systems",
"medium"
] |
J6-CFI-HQ-20K-00020 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | php | refactoring | hard | Refactoring — Pagination Renderer (php) | web_frontend | Refactor the php pagination renderer without changing unrelated behavior. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: eeebc36c7e. | <?php
// case: eeebc36c7e
function normalize_eeebc36c7e(array $items, bool $strict = false): array {
$result = [];
foreach ($items as $raw) {
if ($raw === null) {
if ($strict) {
throw new InvalidArgumentException('Null values are not allowed');
}
conti... | Produce a high-quality refactoring response for a php pagination renderer while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case eeebc36c7e, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a php pagination renderer, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_eeebc36c7e",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_eeebc36c7e",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "eeebc36c7e"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete php pagination renderer.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"php",
"web_frontend",
"hard"
] |
J6-CFI-HQ-20K-00021 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | rust | refactoring | medium | Refactoring — Error Enum (rust) | content_management | Refactor the rust error enum in a content management module. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: 369462a009. | // case: 369462a009
fn normalize_369462a009(items: Vec<Option<String>>, strict: bool) -> Result<Vec<String>, String> {
let mut result = Vec::new();
for raw in items {
match raw {
Some(value) => {
let cleaned = value.trim().to_lowercase();
if !cleaned.is_empty(... | Produce a high-quality refactoring response for a rust error enum while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case 369462a009, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a rust error enum, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_369462a009",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_369462a009",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "369462a009"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete rust error enum.",
"Instruction contains explicit behavioral constraints.",
"Refer... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"rust",
"content_management",
"medium"
] |
J6-CFI-HQ-20K-00022 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | code_review | medium | Code Review — Pagination Builder (python) | authentication | Review the python pagination builder inside a production-facing developer tool. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: cdf8c2f9a0. | # case: cdf8c2f9a0
def normalize_cdf8c2f9a0(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality code review response for a python pagination builder while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case cdf8c2f9a0, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a python pagination builder, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_cdf8c2f9a0",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_cdf8c2f9a0",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "cdf8c2f9a0"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python pagination builder.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"python",
"authentication",
"medium"
] |
J6-CFI-HQ-20K-00023 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | javascript | code_generation | medium | Code Generation — Fetch Retry Wrapper (javascript) | web_frontend | Implement a javascript fetch retry wrapper while preserving an existing workflow. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 0727d6cdf4. | // case: 0727d6cdf4
export function normalize_0727d6cdf4(items, options = { strict: false }) {
const result = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null values are not allowed");
continue;
}
const value = String(raw).trim();
if (value) res... | Produce a high-quality code generation response for a javascript fetch retry wrapper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 0727d6cdf4, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a javascript fetch retry wrapper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_0727d6cdf4",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_0727d6cdf4",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "0727d6cdf4"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete javascript fetch retry wrapper.",
"Instruction contains explicit behavioral constraint... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"javascript",
"web_frontend",
"medium"
] |
J6-CFI-HQ-20K-00024 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | javascript | code_generation | hard | Code Generation — Dom Delegation Helper (javascript) | security | Implement a javascript DOM delegation helper for a realistic service boundary. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: f5f3f089be. | // case: f5f3f089be
export function normalize_f5f3f089be(items, options = { strict: false }) {
const result = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null values are not allowed");
continue;
}
const value = String(raw).trim();
if (value) res... | Produce a high-quality code generation response for a javascript DOM delegation helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case f5f3f089be, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a javascript DOM delegation helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_f5f3f089be",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_f5f3f089be",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "f5f3f089be"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete javascript DOM delegation helper.",
"Instruction contains explicit behavioral constrai... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"javascript",
"security",
"hard"
] |
J6-CFI-HQ-20K-00025 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | sql_data_tasks | medium | Sql Data Tasks — Collection Grouping Utility (java) | security | Improve the SQL collection grouping utility without changing unrelated behavior. Preserve result semantics, handle NULL values, and discuss index or scan implications. Case: 03ed984571. | // case: 03ed984571
public List<String> normalize_03ed984571(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality sql data tasks response for a java collection grouping utility while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Preserve result semantics",
"Handle NULL values deliberately",
"Use explicit joins or filters",
"Discuss index or scan implications"
] | For case 03ed984571, Make joins and filters explicit, define NULL handling, preserve aggregation semantics, and describe the index or query shape that reduces scanning. | This high-quality synthetic sample evaluates whether a coding assistant can handle sql data tasks for a java collection grouping utility, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_03ed984571",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_03ed984571",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "03ed984571"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_join",
"missin... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java collection grouping utility.",
"Instruction contains explicit behavioral constrai... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"sql_data_tasks",
"java",
"security",
"medium"
] |
J6-CFI-HQ-20K-00026 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | code_generation | medium | Code Generation — Repository Query Adapter (java) | data_processing | Implement a java repository query adapter in a data processing module. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: da7290f2ae. | // case: da7290f2ae
public List<String> normalize_da7290f2ae(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality code generation response for a java repository query adapter while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case da7290f2ae, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a java repository query adapter, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_da7290f2ae",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_da7290f2ae",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "da7290f2ae"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java repository query adapter.",
"Instruction contains explicit behavioral constraints... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"java",
"data_processing",
"medium"
] |
J6-CFI-HQ-20K-00027 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | cpp | debugging | hard | Debugging — Topological Sort Helper (cpp) | data_processing | Debug the cpp topological sort helper inside a production-facing developer tool. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: 66011c3a62. | // case: 66011c3a62
vector<string> normalize_66011c3a62(const vector<string>& items) {
vector<string> result;
for (const auto& raw : items) {
string value = trim(raw);
if (!value.empty()) result.push_back(to_lower(value));
}
return result;
} | Produce a high-quality debugging response for a cpp topological sort helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case 66011c3a62, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a cpp topological sort helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_66011c3a62",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_66011c3a62",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "66011c3a62"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete cpp topological sort helper.",
"Instruction contains explicit behavioral constraints."... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"cpp",
"data_processing",
"hard"
] |
J6-CFI-HQ-20K-00028 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | refactoring | hard | Refactoring — Csv Import Pipeline (python) | automation | Refactor the python CSV import pipeline while preserving an existing workflow. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: b8d937a03b. | # case: b8d937a03b
def normalize_b8d937a03b(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality refactoring response for a python CSV import pipeline while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case b8d937a03b, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a python CSV import pipeline, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_b8d937a03b",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_b8d937a03b",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "b8d937a03b"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python CSV import pipeline.",
"Instruction contains explicit behavioral constraints.",... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"python",
"automation",
"hard"
] |
J6-CFI-HQ-20K-00029 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | refactoring | medium | Refactoring — Async Retry Wrapper (python) | database | Refactor the python async retry wrapper for a realistic service boundary. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: a059878870. | # case: a059878870
def normalize_a059878870(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality refactoring response for a python async retry wrapper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case a059878870, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a python async retry wrapper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_a059878870",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_a059878870",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "a059878870"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python async retry wrapper.",
"Instruction contains explicit behavioral constraints.",... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"python",
"database",
"medium"
] |
J6-CFI-HQ-20K-00030 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | code_generation | hard | Code Generation — Pagination Builder (python) | backend_api | Implement a python pagination builder without changing unrelated behavior. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 8ef936ca54. | # case: 8ef936ca54
def normalize_8ef936ca54(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality code generation response for a python pagination builder while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 8ef936ca54, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a python pagination builder, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_8ef936ca54",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_8ef936ca54",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "8ef936ca54"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python pagination builder.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"python",
"backend_api",
"hard"
] |
J6-CFI-HQ-20K-00031 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | refactoring | medium | Refactoring — Pydantic Validator (python) | web_frontend | Refactor the python Pydantic validator in a web frontend module. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: ce962ea037. | # case: ce962ea037
def normalize_ce962ea037(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality refactoring response for a python Pydantic validator while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case ce962ea037, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a python Pydantic validator, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_ce962ea037",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_ce962ea037",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "ce962ea037"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python Pydantic validator.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"python",
"web_frontend",
"medium"
] |
J6-CFI-HQ-20K-00032 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | html_css | algorithmic_reasoning | hard | Algorithmic Reasoning — Navigation Bar (html_css) | cli_tools | Solve the html_css navigation bar problem inside a production-facing developer tool. Explain the invariant, time complexity, space complexity, and boundary behavior. Case: cb08d5d0a3. | /* case: cb08d5d0a3 */
.component-cb08d5d0a3 {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(18rem, 1fr));
gap: 1rem;
}
.component-cb08d5d0a3 .card {
min-width: 0;
padding: 1rem;
} | Produce a high-quality algorithmic reasoning response for a html_css navigation bar while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"State the algorithmic invariant",
"Give time and space complexity",
"Handle empty and boundary inputs",
"Avoid brute force when a better approach is expected"
] | For case cb08d5d0a3, Describe the invariant, choose the simplest algorithm that meets the complexity target, prove boundary behavior, and state time and space complexity. | This high-quality synthetic sample evaluates whether a coding assistant can handle algorithmic reasoning for a html_css navigation bar, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_cb08d5d0a3",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_cb08d5d0a3",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "cb08d5d0a3"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_complexity",
"... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete html_css navigation bar.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"algorithmic_reasoning",
"html_css",
"cli_tools",
"hard"
] |
J6-CFI-HQ-20K-00033 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | javascript | coding_agent_instructions | medium | Coding Agent Instructions — Accessible Dropdown (javascript) | web_frontend | Act as a coding agent for the javascript accessible dropdown while preserving an existing workflow. Do not rewrite the whole file; identify the exact edit location and verification step. Case: 9c8d767cb3. | // case: 9c8d767cb3
export function normalize_9c8d767cb3(items, options = { strict: false }) {
const result = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null values are not allowed");
continue;
}
const value = String(raw).trim();
if (value) res... | Produce a high-quality coding agent instructions response for a javascript accessible dropdown while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not rewrite the whole file",
"Identify the exact edit location",
"Make the smallest safe change",
"Include a verification step"
] | For case 9c8d767cb3, Locate the smallest target block, state the exact insertion or replacement point, avoid full-file rewrites, and include a focused verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle coding agent instructions for a javascript accessible dropdown, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_9c8d767cb3",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_9c8d767cb3",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "9c8d767cb3"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"over_rewrite",
"igno... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete javascript accessible dropdown.",
"Instruction contains explicit behavioral constraint... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"coding_agent_instructions",
"javascript",
"web_frontend",
"medium"
] |
J6-CFI-HQ-20K-00034 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | cpp | code_review | medium | Code Review — Matrix Traversal Function (cpp) | database | Review the cpp matrix traversal function for a realistic service boundary. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: b3ffd04d8d. | // case: b3ffd04d8d
vector<string> normalize_b3ffd04d8d(const vector<string>& items) {
vector<string> result;
for (const auto& raw : items) {
string value = trim(raw);
if (!value.empty()) result.push_back(to_lower(value));
}
return result;
} | Produce a high-quality code review response for a cpp matrix traversal function while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case b3ffd04d8d, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a cpp matrix traversal function, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_b3ffd04d8d",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_b3ffd04d8d",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "b3ffd04d8d"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete cpp matrix traversal function.",
"Instruction contains explicit behavioral constraints... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"cpp",
"database",
"medium"
] |
J6-CFI-HQ-20K-00035 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | php | algorithmic_reasoning | hard | Algorithmic Reasoning — Csrf Validator (php) | cli_tools | Solve the php CSRF validator problem without changing unrelated behavior. Explain the invariant, time complexity, space complexity, and boundary behavior. Case: e1e7cc2df8. | <?php
// case: e1e7cc2df8
function normalize_e1e7cc2df8(array $items, bool $strict = false): array {
$result = [];
foreach ($items as $raw) {
if ($raw === null) {
if ($strict) {
throw new InvalidArgumentException('Null values are not allowed');
}
conti... | Produce a high-quality algorithmic reasoning response for a php CSRF validator while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"State the algorithmic invariant",
"Give time and space complexity",
"Handle empty and boundary inputs",
"Avoid brute force when a better approach is expected"
] | For case e1e7cc2df8, Describe the invariant, choose the simplest algorithm that meets the complexity target, prove boundary behavior, and state time and space complexity. | This high-quality synthetic sample evaluates whether a coding assistant can handle algorithmic reasoning for a php CSRF validator, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_e1e7cc2df8",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_e1e7cc2df8",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "e1e7cc2df8"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_complexity",
"... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete php CSRF validator.",
"Instruction contains explicit behavioral constraints.",
"Re... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"algorithmic_reasoning",
"php",
"cli_tools",
"hard"
] |
J6-CFI-HQ-20K-00036 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | sql | code_generation | hard | Code Generation — Index-Aware Search Query (sql) | testing | Implement a sql index-aware search query in a testing module. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 565bc4e2ae. | -- case: 565bc4e2ae
SELECT account_id,
COUNT(*) AS total_events,
MAX(created_at) AS last_event_at
FROM events_565bc4
WHERE created_at >= CURRENT_DATE - INTERVAL '30 days'
AND account_id IS NOT NULL
GROUP BY account_id
ORDER BY total_events DESC; | Produce a high-quality code generation response for a sql index-aware search query while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 565bc4e2ae, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a sql index-aware search query, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_565bc4e2ae",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_565bc4e2ae",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "565bc4e2ae"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete sql index-aware search query.",
"Instruction contains explicit behavioral constraints.... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"sql",
"testing",
"hard"
] |
J6-CFI-HQ-20K-00037 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | code_generation | easy | Code Generation — Async Retry Wrapper (python) | developer_tools | Implement a python async retry wrapper inside a production-facing developer tool. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 3abf9f0822. | # case: 3abf9f0822
def normalize_3abf9f0822(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality code generation response for a python async retry wrapper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 3abf9f0822, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a python async retry wrapper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_3abf9f0822",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_3abf9f0822",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "3abf9f0822"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python async retry wrapper.",
"Instruction contains explicit behavioral constraints.",... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"python",
"developer_tools",
"easy"
] |
J6-CFI-HQ-20K-00038 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | javascript | algorithmic_reasoning | hard | Algorithmic Reasoning — Service Worker Cache Helper (javascript) | data_processing | Solve the javascript service worker cache helper problem while preserving an existing workflow. Explain the invariant, time complexity, space complexity, and boundary behavior. Case: 0577017366. | // case: 0577017366
export function normalize_0577017366(items, options = { strict: false }) {
const result = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null values are not allowed");
continue;
}
const value = String(raw).trim();
if (value) res... | Produce a high-quality algorithmic reasoning response for a javascript service worker cache helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"State the algorithmic invariant",
"Give time and space complexity",
"Handle empty and boundary inputs",
"Avoid brute force when a better approach is expected"
] | For case 0577017366, Describe the invariant, choose the simplest algorithm that meets the complexity target, prove boundary behavior, and state time and space complexity. | This high-quality synthetic sample evaluates whether a coding assistant can handle algorithmic reasoning for a javascript service worker cache helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_0577017366",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_0577017366",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "0577017366"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_complexity",
"... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete javascript service worker cache helper.",
"Instruction contains explicit behavioral co... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"algorithmic_reasoning",
"javascript",
"data_processing",
"hard"
] |
J6-CFI-HQ-20K-00039 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | debugging | hard | Debugging — Feature Flag Client (typescript) | file_processing | Debug the typescript feature flag client for a realistic service boundary. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: 1e815f5521. | // case: 1e815f5521
type NormalizeOptions_1e815f5521 = { strict?: boolean };
export function normalize_1e815f5521(items: unknown[], options: NormalizeOptions_1e815f5521 = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality debugging response for a typescript feature flag client while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case 1e815f5521, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a typescript feature flag client, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_1e815f5521",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_1e815f5521",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "1e815f5521"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript feature flag client.",
"Instruction contains explicit behavioral constraint... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"typescript",
"file_processing",
"hard"
] |
J6-CFI-HQ-20K-00040 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | javascript | code_review | medium | Code Review — Dom Delegation Helper (javascript) | cli_tools | Review the javascript DOM delegation helper without changing unrelated behavior. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: e863471f0a. | // case: e863471f0a
export function normalize_e863471f0a(items, options = { strict: false }) {
const result = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null values are not allowed");
continue;
}
const value = String(raw).trim();
if (value) res... | Produce a high-quality code review response for a javascript DOM delegation helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case e863471f0a, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a javascript DOM delegation helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_e863471f0a",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_e863471f0a",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "e863471f0a"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete javascript DOM delegation helper.",
"Instruction contains explicit behavioral constrai... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"javascript",
"cli_tools",
"medium"
] |
J6-CFI-HQ-20K-00041 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | sql | debugging | hard | Debugging — Retention Cohort Query (sql) | authentication | Debug the sql retention cohort query in a authentication module. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: b2daf224c6. | -- case: b2daf224c6
SELECT account_id,
COUNT(*) AS total_events,
MAX(created_at) AS last_event_at
FROM events_b2daf2
WHERE created_at >= CURRENT_DATE - INTERVAL '30 days'
AND account_id IS NOT NULL
GROUP BY account_id
ORDER BY total_events DESC; | Produce a high-quality debugging response for a sql retention cohort query while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case b2daf224c6, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a sql retention cohort query, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_b2daf224c6",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_b2daf224c6",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "b2daf224c6"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete sql retention cohort query.",
"Instruction contains explicit behavioral constraints.",... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"sql",
"authentication",
"hard"
] |
J6-CFI-HQ-20K-00042 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | cpp | code_explanation | medium | Code Explanation — Range Query Helper (cpp) | data_processing | Explain the cpp range query helper inside a production-facing developer tool. Describe the data flow, one realistic failure mode, and what should be tested next. Case: f18b6d45ea. | // case: f18b6d45ea
vector<string> normalize_f18b6d45ea(const vector<string>& items) {
vector<string> result;
for (const auto& raw : items) {
string value = trim(raw);
if (!value.empty()) result.push_back(to_lower(value));
}
return result;
} | Produce a high-quality code explanation response for a cpp range query helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Explain what the code does before suggesting changes",
"Call out one realistic failure mode",
"Avoid rewriting the code",
"Keep the explanation concise and technical"
] | For case f18b6d45ea, Walk through input handling, transformation, and output construction without rewriting the code. Mention one realistic failure mode and a focused test. | This high-quality synthetic sample evaluates whether a coding assistant can handle code explanation for a cpp range query helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_f18b6d45ea",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_f18b6d45ea",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "f18b6d45ea"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_explanation",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete cpp range query helper.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_explanation",
"cpp",
"data_processing",
"medium"
] |
J6-CFI-HQ-20K-00043 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | refactoring | medium | Refactoring — Service Response Mapper (typescript) | content_management | Refactor the typescript service response mapper while preserving an existing workflow. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: c18cb188be. | // case: c18cb188be
type NormalizeOptions_c18cb188be = { strict?: boolean };
export function normalize_c18cb188be(items: unknown[], options: NormalizeOptions_c18cb188be = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality refactoring response for a typescript service response mapper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case c18cb188be, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a typescript service response mapper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_c18cb188be",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_c18cb188be",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "c18cb188be"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript service response mapper.",
"Instruction contains explicit behavioral constr... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"typescript",
"content_management",
"medium"
] |
J6-CFI-HQ-20K-00044 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | javascript | algorithmic_reasoning | hard | Algorithmic Reasoning — Localstorage Adapter (javascript) | admin_panels | Solve the javascript localStorage adapter problem for a realistic service boundary. Explain the invariant, time complexity, space complexity, and boundary behavior. Case: 6d9ff11d9c. | // case: 6d9ff11d9c
export function normalize_6d9ff11d9c(items, options = { strict: false }) {
const result = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null values are not allowed");
continue;
}
const value = String(raw).trim();
if (value) res... | Produce a high-quality algorithmic reasoning response for a javascript localStorage adapter while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"State the algorithmic invariant",
"Give time and space complexity",
"Handle empty and boundary inputs",
"Avoid brute force when a better approach is expected"
] | For case 6d9ff11d9c, Describe the invariant, choose the simplest algorithm that meets the complexity target, prove boundary behavior, and state time and space complexity. | This high-quality synthetic sample evaluates whether a coding assistant can handle algorithmic reasoning for a javascript localStorage adapter, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_6d9ff11d9c",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_6d9ff11d9c",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "6d9ff11d9c"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_complexity",
"... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete javascript localStorage adapter.",
"Instruction contains explicit behavioral constrain... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"algorithmic_reasoning",
"javascript",
"admin_panels",
"hard"
] |
J6-CFI-HQ-20K-00045 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | csharp | debugging | hard | Debugging — Background Worker (csharp) | web_frontend | Debug the csharp background worker without changing unrelated behavior. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: 62aa8348aa. | // case: 62aa8348aa
public IEnumerable<string> Normalize_62aa8348aa(IEnumerable<object?> items, bool strict = false)
{
foreach (var raw in items)
{
if (raw is null)
{
if (strict) throw new ArgumentException("Null values are not allowed");
continue;
}
var v... | Produce a high-quality debugging response for a csharp background worker while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case 62aa8348aa, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a csharp background worker, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_62aa8348aa",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_62aa8348aa",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "62aa8348aa"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete csharp background worker.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"csharp",
"web_frontend",
"hard"
] |
J6-CFI-HQ-20K-00046 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | debugging | medium | Debugging — Typed Event Bus (typescript) | automation | Debug the typescript typed event bus in a automation module. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: a0fa26966d. | // case: a0fa26966d
type NormalizeOptions_a0fa26966d = { strict?: boolean };
export function normalize_a0fa26966d(items: unknown[], options: NormalizeOptions_a0fa26966d = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality debugging response for a typescript typed event bus while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case a0fa26966d, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a typescript typed event bus, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_a0fa26966d",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_a0fa26966d",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "a0fa26966d"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript typed event bus.",
"Instruction contains explicit behavioral constraints.",... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"typescript",
"automation",
"medium"
] |
J6-CFI-HQ-20K-00047 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | debugging | hard | Debugging — Next.Js Route Handler (typescript) | database | Debug the typescript Next.js route handler inside a production-facing developer tool. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: 4fce6ea1a2. | // case: 4fce6ea1a2
type NormalizeOptions_4fce6ea1a2 = { strict?: boolean };
export function normalize_4fce6ea1a2(items: unknown[], options: NormalizeOptions_4fce6ea1a2 = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality debugging response for a typescript Next.js route handler while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case 4fce6ea1a2, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a typescript Next.js route handler, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_4fce6ea1a2",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_4fce6ea1a2",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "4fce6ea1a2"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript Next.js route handler.",
"Instruction contains explicit behavioral constrai... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"typescript",
"database",
"hard"
] |
J6-CFI-HQ-20K-00048 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | go | code_review | expert | Code Review — Graceful Shutdown Routine (go) | cli_tools | Review the go graceful shutdown routine while preserving an existing workflow. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: f9ad8e92a4. | // case: f9ad8e92a4
func Normalize_f9ad8e92a4(items []string, strict bool) ([]string, error) {
result := make([]string, 0, len(items))
for _, raw := range items {
value := strings.TrimSpace(raw)
if value == "" {
if strict { return nil, fmt.Errorf("empty values are not allowed") }
... | Produce a high-quality code review response for a go graceful shutdown routine while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case f9ad8e92a4, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a go graceful shutdown routine, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_f9ad8e92a4",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_f9ad8e92a4",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "f9ad8e92a4"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete go graceful shutdown routine.",
"Instruction contains explicit behavioral constraints.... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"go",
"cli_tools",
"expert"
] |
J6-CFI-HQ-20K-00049 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | sql | code_explanation | easy | Code Explanation — Audit Log Aggregation (sql) | web_frontend | Explain the sql audit log aggregation for a realistic service boundary. Describe the data flow, one realistic failure mode, and what should be tested next. Case: 8428287a51. | -- case: 8428287a51
SELECT account_id,
COUNT(*) AS total_events,
MAX(created_at) AS last_event_at
FROM events_842828
WHERE created_at >= CURRENT_DATE - INTERVAL '30 days'
AND account_id IS NOT NULL
GROUP BY account_id
ORDER BY total_events DESC; | Produce a high-quality code explanation response for a sql audit log aggregation while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Explain what the code does before suggesting changes",
"Call out one realistic failure mode",
"Avoid rewriting the code",
"Keep the explanation concise and technical"
] | For case 8428287a51, Walk through input handling, transformation, and output construction without rewriting the code. Mention one realistic failure mode and a focused test. | This high-quality synthetic sample evaluates whether a coding assistant can handle code explanation for a sql audit log aggregation, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_8428287a51",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_8428287a51",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "8428287a51"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_explanation",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete sql audit log aggregation.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_explanation",
"sql",
"web_frontend",
"easy"
] |
J6-CFI-HQ-20K-00050 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | bash | code_generation | medium | Code Generation — Ci Helper (bash) | security | Implement a bash CI helper without changing unrelated behavior. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: c7fd7bf9ae. | #!/usr/bin/env bash
# case: c7fd7bf9ae
set -euo pipefail
input_dir="${1:-./logs}"
grep -R "ERROR" "$input_dir" | awk '{print $1}' | sort | uniq -c | Produce a high-quality code generation response for a bash CI helper while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case c7fd7bf9ae, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a bash CI helper, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_c7fd7bf9ae",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_c7fd7bf9ae",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "c7fd7bf9ae"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete bash CI helper.",
"Instruction contains explicit behavioral constraints.",
"Refere... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"bash",
"security",
"medium"
] |
J6-CFI-HQ-20K-00051 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | sql_data_tasks | medium | Sql Data Tasks — Dictionary Normalizer (python) | data_processing | Improve the SQL dictionary normalizer in a data processing module. Preserve result semantics, handle NULL values, and discuss index or scan implications. Case: 477963b744. | # case: 477963b744
def normalize_477963b744(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality sql data tasks response for a python dictionary normalizer while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Preserve result semantics",
"Handle NULL values deliberately",
"Use explicit joins or filters",
"Discuss index or scan implications"
] | For case 477963b744, Make joins and filters explicit, define NULL handling, preserve aggregation semantics, and describe the index or query shape that reduces scanning. | This high-quality synthetic sample evaluates whether a coding assistant can handle sql data tasks for a python dictionary normalizer, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_477963b744",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_477963b744",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "477963b744"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"wrong_join",
"missin... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python dictionary normalizer.",
"Instruction contains explicit behavioral constraints.... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"sql_data_tasks",
"python",
"data_processing",
"medium"
] |
J6-CFI-HQ-20K-00052 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | php | code_review | medium | Code Review — Upload Validator (php) | web_frontend | Review the php upload validator inside a production-facing developer tool. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: 12a53ddfa7. | <?php
// case: 12a53ddfa7
function normalize_12a53ddfa7(array $items, bool $strict = false): array {
$result = [];
foreach ($items as $raw) {
if ($raw === null) {
if ($strict) {
throw new InvalidArgumentException('Null values are not allowed');
}
conti... | Produce a high-quality code review response for a php upload validator while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case 12a53ddfa7, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a php upload validator, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_12a53ddfa7",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_12a53ddfa7",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "12a53ddfa7"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete php upload validator.",
"Instruction contains explicit behavioral constraints.",
"... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"php",
"web_frontend",
"medium"
] |
J6-CFI-HQ-20K-00053 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | typescript | debugging | hard | Debugging — React Form Component (typescript) | payment_systems | Debug the typescript React form component while preserving an existing workflow. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: b7fd74b43d. | // case: b7fd74b43d
type NormalizeOptions_b7fd74b43d = { strict?: boolean };
export function normalize_b7fd74b43d(items: unknown[], options: NormalizeOptions_b7fd74b43d = {}): string[] {
const result: string[] = [];
for (const raw of items) {
if (raw == null) {
if (options.strict) throw new Error("Null v... | Produce a high-quality debugging response for a typescript React form component while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case b7fd74b43d, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a typescript React form component, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_b7fd74b43d",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_b7fd74b43d",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "b7fd74b43d"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete typescript React form component.",
"Instruction contains explicit behavioral constrain... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"typescript",
"payment_systems",
"hard"
] |
J6-CFI-HQ-20K-00054 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | cpp | debugging | hard | Debugging — String Token Parser (cpp) | payment_systems | Debug the cpp string token parser for a realistic service boundary. Identify the root cause before patching, make the narrowest safe fix, and add a regression check. Case: a75af7900e. | // case: a75af7900e
vector<string> normalize_a75af7900e(const vector<string>& items) {
vector<string> result;
for (const auto& raw : items) {
string value = trim(raw);
if (!value.empty()) result.push_back(to_lower(value));
}
return result;
} | Produce a high-quality debugging response for a cpp string token parser while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Identify the likely root cause before proposing code",
"Make the narrowest safe fix",
"Preserve the public API and current call sites",
"Include one regression check"
] | For case a75af7900e, Start by reproducing the failure path, isolate the invalid assumption, patch the narrowest branch, preserve the contract, and add a regression check that fails before the fix. | This high-quality synthetic sample evaluates whether a coding assistant can handle debugging for a cpp string token parser, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_a75af7900e",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_a75af7900e",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
},
{
... | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "a75af7900e"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"incorrect_root_cause",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete cpp string token parser.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"debugging",
"cpp",
"payment_systems",
"hard"
] |
J6-CFI-HQ-20K-00055 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | go | code_generation | easy | Code Generation — Health-Check Endpoint (go) | content_management | Implement a go health-check endpoint without changing unrelated behavior. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: 8b778d26ce. | // case: 8b778d26ce
func Normalize_8b778d26ce(items []string, strict bool) ([]string, error) {
result := make([]string, 0, len(items))
for _, raw := range items {
value := strings.TrimSpace(raw)
if value == "" {
if strict { return nil, fmt.Errorf("empty values are not allowed") }
... | Produce a high-quality code generation response for a go health-check endpoint while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case 8b778d26ce, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a go health-check endpoint, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_8b778d26ce",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_8b778d26ce",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "8b778d26ce"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete go health-check endpoint.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"go",
"content_management",
"easy"
] |
J6-CFI-HQ-20K-00056 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | rust | code_review | hard | Code Review — File Reader (rust) | backend_api | Review the rust file reader in a backend api module. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: e2f700d4d0. | // case: e2f700d4d0
fn normalize_e2f700d4d0(items: Vec<Option<String>>, strict: bool) -> Result<Vec<String>, String> {
let mut result = Vec::new();
for raw in items {
match raw {
Some(value) => {
let cleaned = value.trim().to_lowercase();
if !cleaned.is_empty(... | Produce a high-quality code review response for a rust file reader while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case e2f700d4d0, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a rust file reader, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_e2f700d4d0",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_e2f700d4d0",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "e2f700d4d0"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete rust file reader.",
"Instruction contains explicit behavioral constraints.",
"Refe... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"rust",
"backend_api",
"hard"
] |
J6-CFI-HQ-20K-00057 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | python | code_generation | hard | Code Generation — Webhook Processor (python) | file_processing | Implement a python webhook processor inside a production-facing developer tool. Handle invalid input, keep the implementation dependency-light, and include a verification example. Case: ab42681f58. | # case: ab42681f58
def normalize_ab42681f58(items, *, strict=False):
result = []
for raw in items:
if raw is None:
if strict:
raise ValueError("None is not allowed")
continue
value = str(raw).strip()
if value:
result.append(value.lower(... | Produce a high-quality code generation response for a python webhook processor while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Keep the implementation dependency-light",
"Handle invalid or boundary input explicitly",
"Return deterministic output",
"Include a small verification example"
] | For case ab42681f58, Create a compact implementation that validates inputs at the boundary, keeps control flow explicit, avoids unnecessary dependencies, and demonstrates normal plus boundary behavior. | This high-quality synthetic sample evaluates whether a coding assistant can handle code generation for a python webhook processor, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_ab42681f58",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_ab42681f58",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": true,
"preserves_api": false,
"risk_level": "low",
"quality_tier": "hq_synthetic",
"unique_case_id": "ab42681f58"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"missing_edge_case",
... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete python webhook processor.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_generation",
"python",
"file_processing",
"hard"
] |
J6-CFI-HQ-20K-00058 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | php | code_review | medium | Code Review — Template Partial (php) | file_processing | Review the php template partial while preserving an existing workflow. Prioritize correctness, data safety, maintainability, and test coverage with actionable findings. Case: 4d1f1f5463. | <?php
// case: 4d1f1f5463
function normalize_4d1f1f5463(array $items, bool $strict = false): array {
$result = [];
foreach ($items as $raw) {
if ($raw === null) {
if ($strict) {
throw new InvalidArgumentException('Null values are not allowed');
}
conti... | Produce a high-quality code review response for a php template partial while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Prioritize correctness and data safety before style",
"Provide actionable findings with severity",
"Avoid vague feedback",
"Mention at least one test or verification step"
] | For case 4d1f1f5463, Return prioritized findings with severity, impact, and remediation. Focus on correctness and safety before style and include at least one verification step. | This high-quality synthetic sample evaluates whether a coding assistant can handle code review for a php template partial, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_4d1f1f5463",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_4d1f1f5463",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": true,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": false,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "4d1f1f5463"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"generic_review",
"mi... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete php template partial.",
"Instruction contains explicit behavioral constraints.",
"... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"code_review",
"php",
"file_processing",
"medium"
] |
J6-CFI-HQ-20K-00059 | J6-CFI-HQ-20K | JumpLander High-Quality 20K Code Feedback Instructions | 1.0.0-hq-synthetic | en | java | refactoring | medium | Refactoring — Authorization Guard (java) | database | Refactor the java authorization guard for a realistic service boundary. Preserve behavior and public names, reduce complexity, and explain why the refactor is safe. Case: 7fd857d8af. | // case: 7fd857d8af
public List<String> normalize_7fd857d8af(List<Object> items, boolean strict) {
List<String> result = new ArrayList<>();
for (Object raw : items) {
if (raw == null) {
if (strict) throw new IllegalArgumentException("Null values are not allowed");
continue;
... | Produce a high-quality refactoring response for a java authorization guard while preserving explicit developer constraints, edge-case behavior, and actionable engineering clarity. | [
"Do not change observable behavior",
"Keep public names stable unless asked otherwise",
"Reduce complexity without broad rewrites",
"Explain the safety of the refactor"
] | For case 7fd857d8af, Extract only the unclear or repeated logic into a focused helper, keep public names stable, compare before and after behavior, and avoid broad rewrites. | This high-quality synthetic sample evaluates whether a coding assistant can handle refactoring for a java authorization guard, preserve constraints, avoid generic output, and provide verifiable engineering guidance. | [
{
"name": "normal_behavior_7fd857d8af",
"input": "representative valid input",
"expected": "documented successful output"
},
{
"name": "boundary_behavior_7fd857d8af",
"input": "empty, missing, minimum-size, or whitespace-only input",
"expected": "safe deterministic handling"
}
] | {
"requires_reasoning": false,
"has_edge_cases": true,
"has_tests": false,
"preserves_api": true,
"risk_level": "medium",
"quality_tier": "hq_synthetic",
"unique_case_id": "7fd857d8af"
} | {
"checks": [
"Matches the developer intent",
"Preserves explicit constraints",
"Avoids unnecessary rewrites or dependencies",
"Includes actionable and testable guidance",
"Handles at least one edge case",
"Avoids generic filler"
],
"failure_types_to_avoid": [
"behavior_change",
"r... | {
"status": "synthetic_hq_candidate",
"duplicate_policy": "exact record, instruction, and code-context duplicates removed during generation",
"human_reviewed": false,
"quality_notes": [
"Scenario is tied to a concrete java authorization guard.",
"Instruction contains explicit behavioral constraints.",
... | [
"coding",
"code-feedback",
"high-quality-synthetic",
"instruction-tuning",
"refactoring",
"java",
"database",
"medium"
] |
End of preview. Expand in Data Studio
J6-CFI-HQ-20K: JumpLander High-Quality 20K Code Feedback Instructions
A high-quality 20K English code instruction dataset for coding assistants, debugging, code review, refactoring, test generation, software engineering reasoning, and coding-agent behavior.
Built by JumpLander for experiments in coding assistants, software engineering datasets, and agentic developer intelligence.
Overview
J6-CFI-HQ-20K stands for JumpLander High-Quality 20K Code Feedback Instructions.
This dataset is an English-only, synthetic, high-quality structured code-instruction dataset designed for supervised fine-tuning experiments, coding assistant prototyping, debugging behavior, code review, refactoring, test generation, algorithmic reasoning, backend/API task design, SQL and data engineering tasks, and coding-agent behavior research.
Unlike minimal instruction -> answer datasets, J6-CFI-HQ-20K includes structured metadata for each record: developer intent, constraints, reference solution, explanation, test cases, quality signals, evaluation checks, and curation metadata.
This release is intentionally smaller than large synthetic corpora. The focus is cleaner structure, stronger constraints, more realistic developer tasks, and better record-level metadata.
Dataset Information
| Property | Value |
|---|---|
| Dataset Name | J6-CFI-HQ-20K |
| Full Name | JumpLander High-Quality 20K Code Feedback Instructions |
| Total Samples | 20,000 |
| Language | English |
| Format | JSONL |
| File | j6_cfi_hq_20k.jsonl |
| License | CC BY 4.0 |
| Data Type | High-Quality Synthetic / Structured |
| Version | 1.0.0-hq-synthetic |
| Created By | JumpLander |
Task Distribution
| Task Type | Samples |
|---|---|
debugging |
4,000 |
code_generation |
3,500 |
code_review |
2,500 |
refactoring |
2,500 |
test_generation |
2,000 |
algorithmic_reasoning |
1,800 |
code_explanation |
1,500 |
sql_data_tasks |
1,000 |
api_backend |
800 |
coding_agent_instructions |
400 |
Programming Language Distribution
| Programming Language | Samples |
|---|---|
python |
5,500 |
javascript |
3,000 |
typescript |
2,000 |
java |
1,700 |
cpp |
1,500 |
sql |
1,400 |
go |
1,000 |
php |
900 |
rust |
800 |
bash |
600 |
html_css |
600 |
csharp |
500 |
ruby |
500 |
Quality Positioning
J6-CFI-HQ-20K is not a human-verified Gold benchmark. It is a high-quality synthetic dataset candidate designed to be cleaner and more focused than large-scale raw synthetic corpora.
Generation and curation properties:
- 20,000 valid JSONL records
- exact duplicate full records: 0
- exact duplicate instructions: 0
- exact duplicate code contexts: 0
- each record includes explicit constraints
- each record includes test cases or validation scenarios
- each record includes quality signals and evaluation checks
- each record includes curation metadata
- designed for SFT and coding-assistant prototyping
For stricter evaluation, a future J6-CFI-Gold-1K subset should be manually reviewed.
Dataset Schema
Each JSONL line contains one record with the following fields:
| Field | Type | Description |
|---|---|---|
id |
string | Unique sample identifier |
dataset |
string | Short dataset name |
full_name |
string | Full dataset name |
version |
string | Dataset version |
language |
string | Natural language of the instruction |
programming_language |
string | Target programming language |
task_type |
string | Main task category |
difficulty |
string | Difficulty level |
title |
string | Short task title |
domain |
string | Software engineering domain |
instruction |
string | User-facing coding instruction |
code_context |
string | Code or project context |
intent |
string | Normalized developer intent |
constraints |
array | Constraints the assistant should preserve |
reference_solution |
string | Reference answer or solution outline |
explanation |
string | Explanation of the expected behavior |
test_cases |
array | Test or validation cases |
quality_signals |
object | Metadata for reasoning, tests, edge cases and risk |
evaluation |
object | Checks for scoring model responses |
curation |
object | Curation and review metadata |
tags |
array | Search and filtering tags |
Loading the Dataset
pip install datasets
from datasets import load_dataset
dataset = load_dataset("jumplander/J6-CFI-HQ-20K")
train = dataset["train"]
print(train[0]["instruction"])
print(train[0]["reference_solution"])
Load Locally
from datasets import load_dataset
dataset = load_dataset(
"json",
data_files={{"train": "j6_cfi_hq_20k.jsonl"}}
)
Formatting for SFT
def format_example(example):
return {{
"text": f"""### Instruction:
{{example['instruction']}}
### Code Context:
{{example['code_context']}}
### Developer Intent:
{{example['intent']}}
### Constraints:
{{example['constraints']}}
### Reference Solution:
{{example['reference_solution']}}
### Explanation:
{{example['explanation']}}
### Verification:
{{example['test_cases']}}"""
}}
formatted = dataset["train"].map(format_example)
Suggested Use Cases
| Use Case | Description |
|---|---|
| SFT Training | Fine-tune coding assistants on structured code instructions |
| Debugging Research | Study bug-fixing and root-cause explanation behavior |
| Code Review Training | Train models to produce prioritized review feedback |
| Refactoring Experiments | Improve behavior-preserving code transformation |
| Test Generation | Teach assistants to generate edge-case-aware tests |
| Coding-Agent Research | Evaluate minimal-edit and safe-patch behavior |
| Prompt Evaluation | Compare coding prompts against structured reference behavior |
| Gold Subset Preparation | Select high-value records for manual review |
Important Quality Notes
J6-CFI-HQ-20K is a high-quality synthetic structured dataset, not a fully human-verified benchmark.
It is suitable for SFT experiments, coding assistant prototyping, structured instruction tuning, behavior-aware code generation research, and preparing a manually reviewed Gold subset.
It should not be described as human-verified, production-certified, a complete correctness benchmark, or a replacement for execution-based evaluation.
Recommended next steps:
- create
J6-CFI-Gold-1K, - manually review technical correctness,
- execute a subset of Python, JavaScript, SQL, and Bash examples,
- score solution usefulness and difficulty,
- publish an evaluation script.
Dataset Philosophy
A strong coding assistant should not only produce code. It should understand developer intent, preserve constraints, avoid unsafe rewrites, explain trade-offs, include tests, and produce actionable engineering output.
J6-CFI-HQ-20K is designed around that principle.
License
This dataset is released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.
Attribution is required when redistributing, adapting, or building upon this dataset.
Citation
@misc{{j6_cfi_hq_20k_2026,
title={{J6-CFI-HQ-20K: JumpLander High-Quality 20K Code Feedback Instructions}},
author={{JumpLander Team}},
year={{2026}},
publisher={{Hugging Face}},
url={{https://huggingface.co/datasets/jumplander/J6-CFI-HQ-20K}}
}}
Built by JumpLander
J6-CFI-HQ-20K is built by JumpLander, a research and engineering initiative focused on coding agents, developer intelligence, AI-assisted programming, and practical software engineering datasets.
- Website: jumplander.org
- Persian Website: jumplander.org/fa
- Documentation: jumplander.org/fa/docs
- About: jumplander.org/fa/about
- Hugging Face: huggingface.co/jumplander
- Downloads last month
- -