# Repository Intelligence [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/nirholas/lyra-intel  
**GitHub Stars:** 29  
**npm Downloads (last month):** 28  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/repository-intelligence

## Description
Analyze repos of any size - security scanning code analysis monorepo support

## Tools
Capabilities this server exposes over MCP:

- **analyze-codebase** — Run comprehensive analysis on a codebase including AST parsing, dependency mapping, complexity metrics, and security scanning
- **search-code** — Semantic code search using ML-powered embeddings to find relevant code snippets
- **get-complexity** — Get complexity metrics (cyclomatic, cognitive, Halstead) for code
- **get-security-issues** — Scan codebase for security vulnerabilities, hardcoded secrets, and compliance issues
- **discovery-scan-github** — Scan GitHub for new MCP crypto tool repositories. Returns a list of discovered repositories with metadata.
- **discovery-analyze-repo** — Analyze a GitHub repository to extract MCP tool definitions, assess quality, and check security.
- **discovery-submit-tool** — Submit analyzed tools from a repository to the Lyra Registry. Requires prior analysis.
- **discovery-run-pipeline** — Run the complete discovery pipeline: scan GitHub, analyze repos, and optionally submit to registry.
- **discovery-get-stats** — Get statistics about the discovery module: recent runs, tools found, submission rates.
- **diff-analyze** — Analyze git diffs between commits/branches to understand changes, their impact, and risk level. Identifies affected functions, classes, and provides actionable recommendations.
- **generate-docs** — Auto-generate documentation from code analysis. Supports API docs, README generation, and architecture documentation in multiple formats.
- **forensic-scan** — Deep forensic analysis to find dead code, technical debt, orphaned documentation, and code-doc discrepancies. Identifies unused functions, classes, and provides actionable cleanup recommendations.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "repository-intelligence": {
    "command": "npx",
    "args": ["-y","lyra-intel-mcp"]
  }
}
```

## Documentation

## What Repository Intelligence does

Repository Intelligence exposes Lyra Intel’s code intelligence functions through MCP. It is designed for projects ranging from small repositories to large monorepos and supports analysis tasks that go beyond simple keyword search or linting.

The core analysis path examines source structure with AST parsing, maps dependencies, calculates code complexity, and scans for security concerns. Semantic search uses ML-powered embeddings to locate relevant snippets by meaning rather than exact text matches. Forensic analysis adds checks for unused functions, classes, technical debt, orphaned documentation, and mismatches between code and documentation.

The server also includes a separate discovery workflow for GitHub. That workflow finds MCP crypto tool repositories, extracts and evaluates their tool definitions, checks security and quality, and can submit analyzed tools to the Lyra Registry.

## How it works

`analyze-codebase` runs a broad repository assessment covering structure, dependencies, complexity, and security. More focused tools expose individual results: `search-code` performs semantic lookup, `get-complexity` returns cyclomatic, cognitive, and Halstead metrics, and `get-security-issues` reports vulnerabilities, hardcoded secrets, and compliance issues.

`diff-analyze` compares commits or branches and evaluates affected functions, classes, impact, and risk. `generate-docs` turns analysis results into API documentation, README content, or architecture documentation in multiple formats. `forensic-scan` looks for dead code, technical debt, unused symbols, and code-documentation discrepancies.

The discovery tools form a pipeline: scan GitHub, analyze selected repositories, optionally submit analyzed tools, and inspect discovery statistics such as recent runs, tools found, and submission rates.

## Setup and configuration

The project supports Python 3.9 or newer and provides Docker and Kubernetes deployment options. Its architecture includes local analysis for development, distributed processing for larger workloads, and cloud-oriented scaling for very large repositories.

Storage options include SQLite for local projects, PostgreSQL for production deployments, and BigQuery for large-scale analysis results. A cache layer can use memory, files, or Redis. AI-assisted analysis can use OpenAI, Anthropic, or local models such as Ollama and llama.cpp.

The broader platform also includes REST API, authentication, OAuth 2.0, LDAP, and role-based access features. These capabilities are relevant when deploying Lyra Intel as a shared service rather than using it only for local analysis.

## Tools and capabilities

- `analyze-codebase`: Run combined codebase analysis.
- `search-code`: Find code semantically with ML-powered embeddings.
- `get-complexity`: Retrieve cyclomatic, cognitive, and Halstead metrics.
- `get-security-issues`: Identify vulnerabilities, secrets, and compliance issues.
- `diff-analyze`: Assess changes between commits or branches.
- `generate-docs`: Produce API, README, and architecture documentation.
- `forensic-scan`: Detect dead code, technical debt, and documentation gaps.
- `discovery-scan-github`: Find MCP crypto tool repositories on GitHub.
- `discovery-analyze-repo`: Extract tool definitions and assess repository quality and security.
- `discovery-submit-tool`: Submit previously analyzed tools to the Lyra Registry.
- `discovery-run-pipeline`: Run the discovery, analysis, and optional submission sequence.
- `discovery-get-stats`: View discovery activity and result statistics.

_Full upstream README: https://allmcps.com/mcp/repository-intelligence/readme_