convez.ai
MCP Server
cognis-digital

repo-roast

cognis-digital/repo-roast

Ver no GitHub
BomPluginScore · de 0 a 100

An AI roasts (and then constructively fixes) your repo — local, free, savage

Instalação

$npx convez-ai cognis-digital/repo-roast

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
100%
Último commit
há 4 meses
Versão
—
Issues abertas
0
Contribuidores
2

README

A tradução para o português deste README ainda está na fila.

repo-roast 🔥

Point an AI at your repo and let it roast you — then hand you 3 concrete fixes. Local, free, savage.

License: COCL 1.0 Local Suite

#developer-tools #ai #fun #code-review #llm

pip install cognis-repo-roast
repo-roast .                 # uses a local model (uncensored-fleet) if running
repo-roast . --no-llm        # heuristic-only roast, no model needed

Usage — step by step

  1. Install the console script (PyPI name cognis-repo-roast, command repo-roast):
    pip install cognis-repo-roast
    
  2. Roast a repo. The single positional argument is a path (defaults to .). With a local model running it uses the LLM; otherwise add --no-llm:
    repo-roast .                 # roast the current repo
    repo-roast ../some-project --no-llm   # heuristic-only, no model
    
  3. Get machine-readable output for scripting — --format json prints the score and roast as JSON:
    repo-roast . --no-llm --format json > roast.json
    
  4. Read the result. The JSON object contains score (0–100) and roast (the text + fixes). Pull the score out, e.g.:
    python -c "import json;print(json.load(open('roast.json'))['score'])"
    
  5. Wire it into CI to track repo hygiene over time:
    - run: pip install cognis-repo-roast
    - run: repo-roast . --no-llm --format json | tee roast.json
    
    Point it at any OpenAI-compatible endpoint via ROAST_ENDPOINT to enable the LLM roast in CI.

Architecture

flowchart LR
  R[Your repo] --> S[Signal scan<br/>README/tests/CI/license/docker]
  S --> H[Heuristic score]
  S --> L[LLM roast + fixes]
  H --> O[Score /100]
  L --> O2[🔥 Roast + 3 fixes]

Use it from any AI stack

Talks to any OpenAI-compatible endpoint (default: uncensored-fleet uncensored slot); set ROAST_ENDPOINT. Works MCP-side too via JSON.

Related

🤖 uncensored-fleet · 📝 readme tooling in the suite

⭐ Star it, then go fix your README.

Interoperability

repo-roast composes with the 300+ tool Cognis suite — JSON in/out and a shared OpenAI-compatible /v1 backbone. See INTEROP.md for the suite map, composition patterns, and reference stacks.

Integrations

Forward repo-roast's findings to STIX/MISP/Sigma/Splunk/Elastic/Slack/webhooks via cognis-connect. See INTEGRATIONS.md.

License

COCL v1.0 — see LICENSE.

Ver o original no GitHub

Bom para

Relacionados