Council Code Review Skill
Get multiple AI perspectives on code changes with structured, actionable feedback.
When to Use
- Review pull requests before merging
- Get code quality feedback on implementations
- Identify potential issues across multiple dimensions
- Validate changes against coding standards
Workflow
- Prepare Input: Provide file paths or git diff
- Invoke Review: Call
mcp:llm-council/verify with code-review rubric
- Process Feedback: Receive structured scores and issue list
- Address Issues: Fix blocking issues before proceeding
Parameters
| Parameter |
Type |
Default |
Description |
snapshot_id |
string |
required |
Git commit SHA for reproducibility |
file_paths |
list |
null |
List of files to review (full file analysis) |
git_diff |
string |
null |
Unified diff format for change-focused review |
rubric_focus |
string |
null |
Focus area: "Security", "Performance", etc. |
tier |
string |
"high" |
Confidence tier: "quick", "balanced", "high", "reasoning" |
Tier Selection Guide
| Tier |
Use When |
Timeout |
balanced |
Routine code reviews |
~90s |
high |
Quality-critical reviews (default) |
~180s |
reasoning |
Complex architectural or security reviews |
~600s |
Input Formats
Supports both:
file_paths: List of files to review (full file analysis)
git_diff: Unified diff format for change-focused review
snapshot_id: Git commit SHA (required for reproducibility)
Rubric (ADR-016)
| Dimension |
Weight |
Focus |
| Accuracy |
35% |
Correctness, no bugs, logic errors |
| Completeness |
20% |
All requirements addressed |
| Clarity |
20% |
Readable, maintainable code |
| Conciseness |
15% |
No unnecessary complexity |
| Relevance |
10% |
Addresses stated requirements |
Output Schema
{
"verdict": "pass|fail|unclear",
"confidence": 0.82,
"rubric_scores": {
"accuracy": 7.5,
"completeness": 8.0,
"clarity": 9.0,
"conciseness": 8.5,
"relevance": 9.0
},
"blocking_issues": [
{
"severity": "major",
"file": "src/api.py",
"line": 42,
"message": "Missing input validation"
}
],
"rationale": "Overall, the code is well-structured...",
"partial": false,
"timeout_fired": false,
"completed_stages": ["stage1", "stage2", "stage3"]
}
Timeout Behavior (ADR-040)
If timeout_fired: true, the review timed out. Check completed_stages to see progress. Consider using a faster tier or reducing the number of files.
Example Usage
# Review specific files
llm-council gate --snapshot abc123 --file-paths src/main.py src/utils.py
# Review git diff
llm-council gate --snapshot $(git rev-parse HEAD) # reviews the snapshot; pass --file-paths to scope
# Review with custom focus
llm-council gate --snapshot $(git rev-parse HEAD) --rubric-focus Security --file-paths src/auth.py
# Deep reasoning review for complex changes
llm-council gate --snapshot $(git rev-parse HEAD) --tier reasoning --rubric-focus Security
Progressive Disclosure
- Level 1: This metadata (~200 tokens)
- Level 2: Full instructions above (~800 tokens)
- Level 3: See
references/code-review-rubric.md for detailed scoring anchors
Related Skills
council-verify: General verification
council-gate: CI/CD quality gate
1---2name: council-review-23description: Multi-model code review with structured feedback using LLM Council peer evaluation. Use for PR reviews, code quality checks, or implementation review. Keywords: code review, PR, pull request, quality check, peer review, feedback4license: Apache-2.05---67# Council Code Review Skill89Get multiple AI perspectives on code changes with structured, actionable feedback.1011## When to Use1213- Review pull requests before merging14- Get code quality feedback on implementations15- Identify potential issues across multiple dimensions16- Validate changes against coding standards1718## Workflow19201. **Prepare Input**: Provide file paths or git diff212. **Invoke Review**: Call `mcp:llm-council/verify` with code-review rubric223. **Process Feedback**: Receive structured scores and issue list234. **Address Issues**: Fix blocking issues before proceeding2425## Parameters2627| Parameter | Type | Default | Description |28|-----------|------|---------|-------------|29| `snapshot_id` | string | required | Git commit SHA for reproducibility |30| `file_paths` | list | null | List of files to review (full file analysis) |31| `git_diff` | string | null | Unified diff format for change-focused review |32| `rubric_focus` | string | null | Focus area: "Security", "Performance", etc. |33| `tier` | string | "high" | Confidence tier: "quick", "balanced", "high", "reasoning" |3435### Tier Selection Guide3637| Tier | Use When | Timeout |38|------|----------|---------|39| `balanced` | Routine code reviews | ~90s |40| `high` | Quality-critical reviews (default) | ~180s |41| `reasoning` | Complex architectural or security reviews | ~600s |4243## Input Formats4445Supports both:46- `file_paths`: List of files to review (full file analysis)47- `git_diff`: Unified diff format for change-focused review48- `snapshot_id`: Git commit SHA (required for reproducibility)4950## Rubric (ADR-016)5152| Dimension | Weight | Focus |53|-----------|--------|-------|54| Accuracy | 35% | Correctness, no bugs, logic errors |55| Completeness | 20% | All requirements addressed |56| Clarity | 20% | Readable, maintainable code |57| Conciseness | 15% | No unnecessary complexity |58| Relevance | 10% | Addresses stated requirements |5960## Output Schema6162```json63{64 "verdict": "pass|fail|unclear",65 "confidence": 0.82,66 "rubric_scores": {67 "accuracy": 7.5,68 "completeness": 8.0,69 "clarity": 9.0,70 "conciseness": 8.5,71 "relevance": 9.072 },73 "blocking_issues": [74 {75 "severity": "major",76 "file": "src/api.py",77 "line": 42,78 "message": "Missing input validation"79 }80 ],81 "rationale": "Overall, the code is well-structured...",82 "partial": false,83 "timeout_fired": false,84 "completed_stages": ["stage1", "stage2", "stage3"]85}86```8788### Timeout Behavior (ADR-040)8990If `timeout_fired: true`, the review timed out. Check `completed_stages` to see progress. Consider using a faster `tier` or reducing the number of files.9192## Example Usage9394```bash95# Review specific files96llm-council gate --snapshot abc123 --file-paths src/main.py src/utils.py9798# Review git diff99llm-council gate --snapshot $(git rev-parse HEAD) # reviews the snapshot; pass --file-paths to scope100101# Review with custom focus102llm-council gate --snapshot $(git rev-parse HEAD) --rubric-focus Security --file-paths src/auth.py103104# Deep reasoning review for complex changes105llm-council gate --snapshot $(git rev-parse HEAD) --tier reasoning --rubric-focus Security106```107108## Progressive Disclosure109110- **Level 1**: This metadata (~200 tokens)111- **Level 2**: Full instructions above (~800 tokens)112- **Level 3**: See `references/code-review-rubric.md` for detailed scoring anchors113114## Related Skills115116- `council-verify`: General verification117- `council-gate`: CI/CD quality gate