Octocode - Agentic Research Platform

Evidence-first code research for AI agents and developers.

Octocode gives an agent the full context it needs to change, review, or explain code: real evidence from your local workspace and from external sources (GitHub repositories, pull requests, and npm packages). One toolset covers all of it: ripgrep and AST structural search, repository tree browsing, precise content fetching, LSP semantic navigation, and binary inspection.

Run it as a CLI or an MCP server. A Rust engine keeps every call fast and token-efficient, minifying and skeletonizing code so an agent reads the shape of a file instead of every byte, from a single file to a mega-repo. It is also the best tool for cross-repository research and exploration across millions of repositories.

Table of Contents

• Why Octocode
• What You Can Do
• Tools
• MCP
• CLI
• Configuration
• Authentication Methods
• Security
• Language Support
• Skills
• Architecture
• Documentation

Why Octocode

Agents write better code from evidence than from assumptions. Octocode turns guess-driven work into research-driven work. Before an agent changes, reviews, or explains code, it gathers real evidence from your local workspace and from GitHub repositories, pull requests, and npm packages, then hands it back as compact, citable context. Code is truth; context is the map.

Most tools cover one slice: searching the web, or grepping your repo. Octocode covers the whole research flow, end to end:

• Built for scale. In organizations with thousands of repositories and endless code, Octocode is the solution: spot a pattern in one repo, follow it through the pull request that introduced it, then trace the same shape across other repos and your own files without leaving the conversation. Clone any repo and study it locally, on any machine.
• Smart GitHub workflows. Parallel bulk queries and built-in next-step hints keep the agent on the cheapest path: search broadly, read narrowly, trace semantically. Each result points to the natural follow-up.
• No GitHub required. Even without GitHub, clone any repository locally and point Octocode's local tools (search, structural AST, LSP, content) at it for the same evidence-first research.
• Reads the shape, not the noise. Code is minified and skeletonized on the fly across 70+ languages, so an agent grasps a 100 KB file in a few hundred tokens instead of spending its context on boilerplate.
• Fast and self-contained. Search, parsing, semantic navigation, and redaction run in one prebuilt Rust engine: quick on a laptop or a mega-repo, with no extra toolchain to install.
• Safe by default. Every byte returned to the model is scanned and secrets redacted first (see Security).

What You Can Do

Octocode is useful whenever the next coding step depends on finding and proving context, not guessing it.

Get Started

Add Octocode to an AI assistant with MCP, or run the same tools directly from your terminal with the CLI.

MCP fast install:

<img src="https://cursor.com/deeplink/mcp-install-dark.svg" alt="Install in Cursor"> <img src="https://img.shields.io/badge/VS_Code-VS_Code?style=flat-square&label=Install%20Server&color=0098FF" alt="Install in VS Code"> <img alt="Install in VS Code Insiders" src="https://img.shields.io/badge/VS_Code_Insiders-VS_Code_Insiders?style=flat-square&label=Install%20Server&color=24bfa5">

# Interactive installer for Cursor, Claude Code, Codex, VS Code, and more
npx octocode install

CLI fast install:

npx octocode

Authenticate GitHub when you want private repositories or higher API limits:

npx octocode login

Benchmarks

Latest benchmark output: packages/octocode-benchmark/output.

ast-grep Structural Comparison

Bars show relative throughput. Higher is better; lower ms is better.

What we tested: ast-grep CLI and Octocode structural grep on the same real repository files, using the same broad AST node-kind searches (call_expression, call, method_invocation). The goal was to check structural AST grep compatibility by match count, then measure where time is spent across Octocode's raw matcher, agent tool path, and public CLI.

This benchmark does not test text grep, LSP navigation, rewriting, or the full ast-grep rule language. Those are separate capabilities.

Octocode raw native  ████████████████████   17.1 ms median  │  2.0x faster  │  6/6 matched
ast-grep CLI         ██████████░░░░░░░░░░   34.6 ms median  │  baseline     │  6/6 matched

Octocode raw native means the direct Rust/NAPI structuralSearchFiles matcher: parse and match only, with no tool validation, sanitizer, pagination, JSON shaping, or Node CLI startup. The agent-facing localSearchCode and public octocode search --pattern/--rule paths are intentionally slower because they include those safety and DX layers.

What was checked: we took ast-grep's benchmark scenario repo list, picked one deterministic file from each supported repo, asked both tools to find the same AST node kind in that file, verified identical match counts, then timed the median run.

Benchmark files: runner · scenario manifest · latest output · unified CLI/tool/OQL eval · agent runbook

Tools

Octocode ships 14 research tools; the same implementations run identically over MCP and the CLI. ghCloneRepo is opt-in for MCP (ENABLE_CLONE=true) and enabled by default for CLI; local tools require ENABLE_LOCAL (CLI default: on, MCP default: off). All flags: Configuration Reference.

Token knobs. concise:true returns path/title-only lists. minify controls file read density: symbols = skeleton with line numbers, standard = comments/blanks stripped (default), none = exact bytes.

GitHub Tools

Local Tools

Package Search

LSP

OQL Search

Per-tool references (full schemas, fields, and examples) live in docs/mcp:

• GitHub Tools
• Local Tools
• Binary Tools
• LSP Tools
• Tool Behavior Guide

MCP

The MCP server exposes the Octocode tool catalog directly to your AI assistant over stdio. Install once; the assistant calls tools automatically.

Install

Fast install:

<img src="https://cursor.com/deeplink/mcp-install-dark.svg" alt="Install in Cursor"> <img src="https://img.shields.io/badge/VS_Code-VS_Code?style=flat-square&label=Install%20Server&color=0098FF" alt="Install in VS Code"> <img alt="Install in VS Code Insiders" src="https://img.shields.io/badge/VS_Code_Insiders-VS_Code_Insiders?style=flat-square&label=Install%20Server&color=24bfa5">

Or use the installer (detects your installed clients):

# Interactive: detects your installed clients
npx octocode install

# Non-interactive
octocode install --ide cursor
octocode install --ide claude-code

https://github.com/user-attachments/assets/de8d14c0-2ead-46ed-895e-09144c9b5071

Manual Configuration

Add to your MCP client config file:

{
  "mcpServers": {
    "octocode": {
      "command": "npx",
      "args": ["octocode-mcp@latest"]
    }
  }
}

For GitHub auth, add a token under env (see Authentication Methods).

Configuration

Set tokens and options as env entries here, or machine-wide in .octocoderc. See Configuration for every setting, the home-folder layout, and precedence.

CLI

The CLI exposes the same research engine without an MCP client. Use quick commands for humans, or call raw tools from scripts and CI. It is agent-friendly by design: npx octocode --help, npx octocode context, and npx octocode tools <name> --scheme publish the research protocol, tool descriptions, and exact schemas, while command output returns compact anchors, pagination, and follow-up hints that guide agents through evidence-first research.

Install

npx octocode

npx octocode login
npx octocode status

All Commands

Local paths route to local tools; owner/repo[/path] targets route to GitHub tools.

Full command syntax, flags, examples, and exit codes live in the CLI Reference.

Configuration

Everything is optional; Octocode runs on sensible defaults. Settings resolve from three sources, in priority order:

environment variables  >  <octocode-home>/.octocoderc  >  built-in defaults

1. MCP / environment variables (highest): per client or per project, set in your MCP config env or your shell.
2. Global config: <octocode-home>/.octocoderc, machine-wide defaults read by both the CLI and the MCP server.
3. Built-in defaults: used when neither is set.

Octocode home (<octocode-home>) holds the global config, encrypted credentials, sessions, stats, and tmp materialization caches. Its location is fixed per platform (there is no override):

Set in MCP (env entries; these win over .octocoderc):

{
  "mcpServers": {
    "octocode": {
      "command": "npx",
      "args": ["octocode-mcp@latest"],
      "env": {
        "GITHUB_TOKEN": "ghp_xxxxxxxx",
        "ENABLE_LOCAL": "true",
        "ENABLE_CLONE": "false"
      }
    }
  }
}

Set globally for both the CLI and MCP in <octocode-home>/.octocoderc (JSON, comments and trailing commas tolerated; never put tokens here). See the ready-to-copy example below.

Common settings

The Scope column shows where a setting applies: Both, or MCP (the CLI ignores it).

Local and clone defaults differ by surface. The CLI honors ENABLE_LOCAL and defaults local tools on; clone is enabled by default. The MCP server honors ENABLE_LOCAL and defaults local tools and clone off, so a deployment can control what an assistant may touch. Explicit env or .octocoderc values override those defaults.

Example .octocoderc

Drop this at <octocode-home>/.octocoderc for machine-wide defaults shared by the CLI and the MCP server. Every field is optional; keep only what you want to change. Tokens never go here (use env or octocode login).

{
  // GitHub Enterprise users: point at your API endpoint.
  "github": { "apiUrl": "https://api.github.com" },

  "local": {
    "enabled": true,            // overrides the surface default for CLI and MCP
    "enableClone": false,       // false disables ghCloneRepo for CLI and MCP
    "workspaceRoot": "~/code",  // base for relative local paths (absolute or ~)
    "allowedPaths": []          // extra dirs the local tools may read
  },

  "network": { "timeout": 30000, "maxRetries": 3 },

  "output": { "format": "yaml" }  // "yaml" or "json"
}

Per-project overrides and custom LSP servers live in a workspace .octocode/ folder (for example .octocode/lsp-servers.json). For every variable, the full .octocoderc schema, clone-cache tuning, GitHub Enterprise setup, local-state paths, and precedence details, see the Configuration Reference.

Authentication Methods

GitHub-backed tools require authentication. Any one method is enough. Full details: Authentication Setup.

Option 1: Octocode CLI (Recommended)

npx octocode login
npx octocode status       # verify the active token source

Interactive login lets you choose Octocode browser OAuth or gh auth login. Octocode OAuth credentials are stored encrypted on disk.

Option 2: GitHub CLI (also supported)

gh auth login

Octocode reads the gh token automatically — no further config needed.

Option 3: Personal Access Token (also supported)

Set OCTOCODE_TOKEN, GH_TOKEN, or GITHUB_TOKEN in your shell. Required scopes: repo, read:user, read:org.

Create a token at github.com/settings/tokens.

Security tip: Never commit tokens to version control. Use environment variables or secure secret management.

Security

Every byte that reaches the model is scanned and redacted first. All content (local files, GitHub and npm responses, error messages, and tool outputs) passes through the Rust engine's secret scanner on the way in (tool inputs) and on the way out (results), so secrets never reach the LLM. The same enforcement runs identically under MCP and the CLI.

• Secret redaction, in and out. 300+ provider credential patterns (AWS, Azure, GCP, GitHub, OpenAI, Anthropic, Stripe, Slack, 1Password, and more) plus generic JWTs, PEM/private keys, bearer tokens, database connection strings, and high-entropy strings. Masked values surface a redaction warning so the agent knows.
• Content sanitized at the source. Local reads (localGetFileContent, ripgrep, structural search, binary, file discovery, structure) and external fetches (GitHub code/files, npm) are scanned as they are read, not only at the boundary.
• Path safety. Relative inputs resolve from WORKSPACE_ROOT / config / cwd, then local reads are bounded to the engine's allowed roots (home by default, plus ALLOWED_PATHS and Octocode-registered roots). Symlinks are resolved and the real target is re-validated, so a link cannot escape into a blocked location.
• Sensitive files and directories are blocked by default. Octocode refuses to read known secret-bearing files and folders wherever they live, returning a redacted error instead of contents. Blocked patterns include:
  • Keys and certs: *.pem, *.key, *.crt/*.cer/*.csr, *.p12/*.pfx/*.jks/*.keystore, and SSH keys (id_rsa, *_ed25519, authorized_keys, known_hosts, .ssh/).
  • Credentials and tokens: .env / .env.*, .netrc, .npmrc, .pgpass, .git-credentials, *_token / .token, client_secret*.json, *service-account*.json, auth.json, .htpasswd.
  • Cloud and infra: .aws/, .azure/, .config/gcloud/, .kube/ / kubeconfig, .docker/, .terraform/ and *.tfstate.
  • OS and app secret stores: .git/, secrets/, private/, browser login data (Chrome/Firefox), OS keychains, password managers (*.kdbx), shell history files, and crypto wallets.
• Command safety. Normal local search runs in-process inside octocode-engine. External helpers are fixed per lane, command/argument allowlisted, and run via spawn with argument arrays: no shell strings, no injection.
• Schema validation runs before any tool executes; untrusted input size and shape are bounded.
• Credentials. GitHub auth via env tokens, AES-256-GCM-encrypted on-disk OAuth, or the gh CLI; tokens are never logged.

Full security model, pipeline, and threat coverage: SECURITY.md. Related: Authentication · Configuration · Credentials

Language Support

Four code-intelligence axes; three are native to the Rust engine and need no external tooling:

📋 Full support matrix: every extension with its exact AST, signature, LSP, and minify capability, machine-generated from the shipped binary, lives in benchmark/SUPPORT.md (150 extensions: 61 AST, 47 signature, 56 LSP, 89 minify-only). Regenerate or verify with yarn workspace @octocodeai/octocode-benchmark matrix:check.

Skills

Agent Skills are a lightweight, open format for extending AI agent capabilities. Browse and install on skills.sh/bgauryy/octocode-mcp · Skills index: skills/README.md

These are the skills the Octocode team itself uses to build Octocode. 9 skills live under skills/; the table mirrors the Skills Index. ⭐ Engineer is the recommended starting skill.

Install any GitHub skill folder directly from the Octocode CLI:

# Shared cross-agent location: ~/.agents/skills
npx octocode skill --add bgauryy/octocode-mcp/skills/octocode-engineer --platform common

# Agent-safe install for multiple clients; never prompts
npx octocode skill --add https://github.com/bgauryy/octocode-mcp/tree/main/skills/octocode-engineer --platform cursor,codex --mode copy --json

Platforms: common (~/.agents/skills), cursor, claude (Claude Code + Claude Desktop), codex, or all. Use --mode symlink to keep each agent pointed at the Octocode-managed source cache. Full guide: Skills Guide.

Architecture

A yarn-workspaces monorepo. The MCP server and the CLI are thin front-ends over one shared TypeScript tool core, which delegates every CPU-heavy path to a single Rust engine (compiled via napi-rs to prebuilt .node binaries). One tool catalog, one security layer, one response shaper, reached two ways.

graph LR
    CLI["octocode<br/>CLI"]
    MCP["octocode-mcp<br/>MCP server, stdio"]
    VSC["VS Code extension<br/>OAuth + install"]
    CORE["octocode-tools-core<br/>tools, GitHub client, auth, pagination, security bridge"]
    ENGINE["octocode-engine (Rust)<br/>secrets, minify, AST, signatures, ripgrep/diff/YAML, LSP"]
    EXT["GitHub API, local FS + ripgrep, language servers"]

    CLI --> CORE
    MCP --> CORE
    VSC -. starts .-> MCP
    CORE --> ENGINE
    CORE --> EXT
    ENGINE --> EXT

    style ENGINE fill:#1a1a2e,stroke:#e75d2a,color:#fff

Request flow is identical whether a call arrives over MCP or the CLI:

client → sanitize inputs (Rust) → run tool (GitHub / FS / LSP) → sanitize + YAML-serialize + paginate (Rust) → result + next-step hints

One Rust engine owns secret detection, sanitization, path/command validation, minification (70+ languages), signature extraction, structural AST search, ripgrep parsing, diff filtering, YAML serialization, and LSP, so the Node event loop stays unblocked and there is no duplicate native loader. It ships prebuilt for darwin (arm64/x64), linux (arm64/x64, gnu + musl), and win32-x64; no Rust toolchain is needed at runtime.

Packages

Documentation

Website: octocode.ai · Product docs: github.com/bgauryy/octocode/tree/main/docs · Index: docs/README.md. Product documentation lives in docs/; benchmark methodology, evals, and run artifacts live in packages/octocode-benchmark.

Docs map

• docs/mcp/: MCP server configuration, authentication, tools, workflows, architecture
• docs/cli/: CLI commands, flags, and reference material
• docs/: guides for development, security, and Pi setup
• packages/octocode-benchmark/: benchmark methodology, support matrix, unified eval, recipes, output schema, and run artifacts

Setup

• Authentication Setup
• Configuration Reference
• Using octocode-mcp with Pi

Tool References

• GitHub Tools
• Local Tools
• Binary Tools
• LSP Tools
• Clone & Local Workflow
• Tool Behavior Guide

Benchmarks & Evals

• Benchmark Summary
• Unified CLI/Tool/OQL Eval
• Benchmark Runbook
• Support Matrix

Security, CLI & Skills

• Security Model
• CLI Reference
• Skills Guide
• Skills Index

Shared Internals

• Credentials Architecture · Session Persistence

Recommended dev mode: Pi + Octocode

Pi is a fast, local-first coding agent whose stated philosophy is "CLI tools with READMEs (Skills) over MCP." Pairing it with Octocode gives a lean, evidence-driven dev loop — Pi edits, Octocode researches. Two routes, pick by how much surface you need:

• Skill route — recommended, leanest. Drop the octocode-engineer skill into Pi's global skills dir. It drives the Octocode CLI directly — no MCP transport, minimal token overhead — and Pi auto-discovers it:

  bash

  npx octocode skill --add bgauryy/octocode-mcp/skills/octocode-engineer --platform common
  

• Adapter route — full tool surface. Install pi-mcp-adapter to expose all 14 Octocode MCP tools behind a single ~200-token proxy tool, so servers stay disconnected until a tool is actually called. Enable clone tools with ENABLE_CLONE=true.

Tune Pi's behavior with an APPEND_SYSTEM.md (a compact starter lives at docs/PI/APPEND_SYSTEM.md). The full walkthrough — adapter install, MCP config scopes, skills, system-prompt tuning, and custom models — is in the Pi Setup Guide.

The Manifest

"Code is Truth, but Context is the Map." Read the Manifest of Octocode for Research Driven Development to understand the philosophy behind Octocode.