| name | commit-impact-analysis |
| description | Automatically analyze git commits to predict which tests might fail and identify potential risks. Use this skill when the user pushes code, creates a PR, or asks about the impact of their changes. |
Commit Impact Analysis
This skill analyzes code changes to predict:
- Which tests are likely to fail
- Security vulnerabilities introduced
- Performance impacts
- Risk score for the change
When to Use
- User runs
git push or git commit
- User asks "what might this break?"
- User creates or updates a PR
- User asks about test coverage for changes
How It Works
- Extract the git diff using the helper script
- Identify changed files and affected components
- Query the Argus API for:
- Test impact graph (file → test mapping)
- Historical failure patterns
- Security scan results
- Generate predictions and recommendations
Implementation
Use the following Argus MCP tools:
-
argus_what_to_test - Analyzes git diff to identify which tests are affected
- Input:
{ "diff": "<git diff output>", "files": ["file1.py", "file2.js"] }
- Returns: List of tests to run with priority and reasoning
-
argus_risk_scores - Calculates risk score for the changes
- Input:
{ "commit": "<commit SHA or branch>" }
- Returns: Risk assessment with scores and deployment recommendation
The tools return:
- Predicted test failures with confidence scores
- Security vulnerabilities found
- Risk assessment (low/medium/high)
- Recommended actions
Output Format
Present findings as a structured report:
- 🧪 Test Predictions (table with test name, failure probability, reason)
- 🔒 Security Findings (severity, location, recommendation)
- 📊 Risk Score with deployment recommendation
- 💡 Suggested Actions