# sqry - Semantic Code Search

AST-backed semantic code search that builds symbol and relation graphs locally. It lets AI assistants trace call paths and analyze dependencies without external requests.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 31
- Author: verivus-oss
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/verivus-oss/sqry
- Updated: 2026-10-06
- HTML: https://affpapa.org/en/nevibecode/mcp/verivus-oss-sqry-semantic-code-search

## What it does

sqry turns source code into AST-derived graphs of symbols and relationships on your local drive. The project ships 37 language plugins, where 28 offer full relationship tracking. Everything runs offline. No telemetry is collected.

The server inspects execution paths between functions and detects cyclic dependencies. It maps cross-language FFI boundaries across languages such as Rust and C. Its shape matching feature discovers functions with matching structural bodies even when identifiers differ.

The standalone sqry-mcp binary provides 39 tools, while the daemon-backed mode exposes 17. Running the server via the packaged setup command requires Docker to pull and execute the container image.

## Requirements
- Needs: Docker

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add verivus-oss-sqry-semantic-code-search -- docker run -i --rm ghcr.io/verivus-oss/sqry:4.8.16
```

### Codex

**Platform CLI**
```bash
codex mcp add verivus-oss-sqry-semantic-code-search -- docker run -i --rm ghcr.io/verivus-oss/sqry:4.8.16
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "verivus-oss-sqry-semantic-code-search": {
            "command": "docker",
            "args": [
                "run",
                "-i",
                "--rm",
                "ghcr.io/verivus-oss/sqry:4.8.16"
            ]
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "verivus-oss-sqry-semantic-code-search": {
            "type": "stdio",
            "command": "docker",
            "args": [
                "run",
                "-i",
                "--rm",
                "ghcr.io/verivus-oss/sqry:4.8.16"
            ]
        }
    }
}
```

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add verivus-oss-sqry-semantic-code-search docker run -i --rm ghcr.io/verivus-oss/sqry:4.8.16
```

## When to use it
- trace call paths between two functions in the codebase
- detect unused symbols and circular dependencies
- inspect cross-language FFI call edges across modules
- generate an architectural overview of repository hotspots

## Example prompts
- Trace the execution path from handle_request to verify_token.
- Run a repository overview with sqry to show complexity hotspots and main hubs.
- Find all functions that have structural shape matches with process_item.

## Similar
- [grepai-languages](https://affpapa.org/en/nevibecode/skills/grepai-languages.md) — skill · 704 installs
- [Paparats – Local Semantic Code Search MCP](https://affpapa.org/en/nevibecode/mcp/paparats-local-semantic-code-search-mcp.md) — MCP server · ★ 11
- [tldr-deep](https://affpapa.org/en/nevibecode/skills/tldr-deep.md) — skill · 481 installs
- [lynx](https://affpapa.org/en/nevibecode/mcp/lynx.md) — MCP server · ★ 10
- [jCodemunch MCP](https://affpapa.org/en/nevibecode/mcp/jcodemunch-mcp.md) — MCP server · ★ 2809
- [Octocode](https://affpapa.org/en/nevibecode/mcp/octocode.md) — MCP server · ★ 518

## Other versions with the same name
- [sqry - Semantic Code Search · verivus-oss](https://affpapa.org/en/nevibecode/mcp/sqry-semantic-code-search.md) — ★ 31

From the same set: [verivus-oss/sqry](https://affpapa.org/en/nevibecode/by/verivus-oss/sqry.md) — 2 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md