Home › Version Control › crg

crg

Token-efficient code review knowledge graph: semantic search and call-graph resolution.

Python

★ 69 stars Last pushed 2026-09-23 License: Apache-2.0

View repository →

Connect

Review any command before running it. Package names and URLs come from the server's own registry entry.

Package (pypi 3.26.23)

uvx better-code-review-graph

Or add to your MCP client config:

{
  "mcpServers": {
    "crg": {
      "command": "uvx",
      "args": [
        "better-code-review-graph"
      ],
      "env": {
        "EMBEDDING_MODELS": "<YOUR_VALUE>",
        "LOCAL_EMBEDDING_MODEL": "<YOUR_VALUE>",
        "LOCAL_RERANK_MODEL": "<YOUR_VALUE>",
        "LOCAL_EMBEDDING_DIM": "<YOUR_VALUE>",
        "LOCAL_EMBEDDING_MODEL_FILE": "<YOUR_VALUE>",
        "LOCAL_EMBEDDING_POOLING": "<YOUR_VALUE>",
        "LOCAL_EMBEDDING_NORMALIZE": "<YOUR_VALUE>",
        "SUMMARY_MODELS": "<YOUR_VALUE>",
        "COHERE_API_KEY": "<YOUR_VALUE>",
        "OPENROUTER_API_KEY": "<YOUR_VALUE>",
        "EMBEDDING_BACKEND": "<YOUR_VALUE>",
        "EMBEDDING_API_BASE": "<YOUR_VALUE>",
        "LLM_API_BASE": "<YOUR_VALUE>",
        "SUMMARY_MODEL": "<YOUR_VALUE>"
      }
    }
  }
}

EMBEDDING_MODELS — Embedding models in selection order; the current runtime uses the first provider/model entry; empty = local ONNX
LOCAL_EMBEDDING_MODEL — Built-in fastretrieval model ID or local directory with fastretrieval-manifest.json
LOCAL_RERANK_MODEL — Fastretrieval TextCrossEncoder model ID; empty = reranking disabled
LOCAL_EMBEDDING_DIM — Required positive output dimension for an external local embedding ID without a manifest
LOCAL_EMBEDDING_MODEL_FILE — ONNX file path inside a manifest-backed local embedding directory
LOCAL_EMBEDDING_POOLING — Explicit pooling for an external local embedding ID without a manifest
LOCAL_EMBEDDING_NORMALIZE — Explicit L2 normalization for an external local embedding ID without a manifest
SUMMARY_MODELS — Summarizer model selection; first provider/model entry only, no fallback; empty = disabled
COHERE_API_KEY secret — Credential for explicitly selected Cohere embeddings, including a configured CF AI Gateway route
OPENROUTER_API_KEY secret — Credential for explicitly selected OpenRouter completion models
EMBEDDING_BACKEND — DEPRECATED (honored one release): embedding backend override 'local' or 'cloud'. Use EMBEDDING_MODELS instead.
EMBEDDING_API_BASE — Provider-compatible cloud embedding endpoint, including CF AI Gateway; SSRF-guarded.
LLM_API_BASE — Provider-compatible summarizer base URL, including CF AI Gateway; SSRF-guarded.
SUMMARY_MODEL — DEPRECATED (honored one release): single summarizer model as a litellm 'provider/model' string. Use SUMMARY_MODELS instead.