carrycash/NeuroDiff — AI Code Review
Semantic git diff analysis for AI-generated code. Detects security issues, architectural violations, and code duplication via AST analysis and optional LLM deep reasoning.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| base-ref | Base Git reference (e.g. HEAD~1 or the base branch SHA) | no | HEAD~1 |
| head-ref | Head Git reference (e.g. HEAD or current commit SHA) | no | HEAD |
| arch-rules | Path to arch-rules.yaml file (optional) | no | "" |
| use-llm | Enable Phase 4 LLM deep analysis (requires API key) | no | false |
| llm-provider | LLM provider to use: gemini | claude | ollama (auto-detected if not set) | no | "" |
| strict | Fail on HIGH severity findings (default: fail only on CRITICAL) | no | false |
| sarif-output | Path to write SARIF v2.1.0 output (for GitHub Security tab) | no | neurodiff.sarif |
| post-pr-comment | Post analysis summary as a PR comment | no | true |
| python-version | Python version to use | no | 3.11 |
Outputs
| name | description |
|---|---|
| sarif-file | Path to the generated SARIF file |
| is-safe | true if the diff is safe to merge, false otherwise |