AI-powered code review CLI for technical hiring.

Analyzes a candidate's Git repository and generates a structured report.md evaluating software engineering foundations and AI orchestration skills across three pillars:

1. Architecture Awareness — Project structure, separation of concerns, design patterns
2. Code Refinement — Code cleanliness, idiomatic usage, absence of boilerplate
3. Edge Case Coverage — Error handling, test coverage, security considerations

Requires Python 3.12+ and uv.

git clone https://github.com/buildfaster-dev/vetter-cli.git
cd vetter-cli
uv sync

Vetter analyzes local repositories only. Clone the candidate's repo first, then point Vetter at it.

export ANTHROPIC_API_KEY=your-key-here

# Clone and analyze
git clone https://github.com/candidate/repo.git
uv run vetter analyze ./repo

# Or analyze a repo already on disk
uv run vetter analyze /path/to/candidate/repo

--candidate and --repo-url are metadata that appear in the report header. They do not affect analysis.

Static analysis that objectively measures:

• Test coverage ratio
• Linter/formatter configuration
• Commit history quality and cadence
• Dependency audit
• Error handling patterns (strategic vs. blanket)
• Security scan (hardcoded secrets)

Sends the codebase to Claude for expert evaluation. Scores each pillar (1-5) with written justification and code evidence.

Combines both layers into a report.md with:

• Classification: Copy-Paster / Assisted Engineer / AI Orchestrator
• Recommendation: Reject / Review Further / Pass
• Pillar scores with justification
• Metrics summary

## Classification

| Metric | Value |
|--------|-------|
| Average Pillar Score | 4.0 / 5 |
| Classification | AI Orchestrator |
| Recommendation | Pass |

# Install dependencies
uv sync

# Run tests
uv run pytest -v

# Run the CLI
uv run vetter --help

MIT