Mighty
AgentsSkillsMCP
⌘K
Open Mighty
Back to MCP
✦

Octocode: Research Driven Development for AI

CommunityDeveloper1.0.0·2026-05-28T04:54:51.163Z014
mcp.details.summary.title

The Octocode MCP Server is a research-driven development environment that connects your AI assistant to code repositories and local tools. You can:

  • Code Search and Exploration: Search GitHub, GitLab, Bitbucket, and local files for code patterns, functions, classes, and symbols using advanced search, file content retrieval, and repository structure visualization.
  • Repository and Package Management: Discover repositories by topics, owners, or keywords; explore package sources from npm, PyPI, and other ecosystems; and examine dependencies and repository structures.
  • Research and Planning: Conduct research on code, bugs, features, or large refactors; plan development tasks with research-backed insights; and review pull requests comprehensively.
  • Local File System Operations: Investigate local files and directories, find files by metadata, read file contents, and navigate codebases for detailed analysis.
  • LLM-Driven Project Scaffolding: Generate new projects with architectural guidance, leveraging research tools for best practices and feature implementation.

Octocode: Research Driven Development for AI

Octocode Logo

Stop Guessing. Start Knowing.

Empower your AI assistant with the skills of a Senior Staff Engineer.

octocode.ai

Installation

Prerequisites: GitHub authentication required. See Authentication Setup.

npx octocode-cli install

Interactive setup wizard with GitHub OAuth, MCP server installation, and skills marketplace. Pass --ide <ide> for non-interactive install (e.g. npx octocode-cli install --ide cursor).

Alternative Methods

One-Click Install (Cursor)

Install in Cursor

Manual MCP Configuration

Add to your MCP configuration file:

JSON
{
  "mcpServers": {
    "octocode": {
      "command": "npx",
      "args": ["octocode-mcp@latest"]
    }
  }
}
Research Skill (Direct Install)
Shell
npx add-skill https://github.com/bgauryy/octocode-mcp/tree/main/skills/octocode-research

MCP Server

The Octocode MCP Server connects your AI assistant to code:

  • GitHub, GitLab & Bitbucket: Search repositories, find usage patterns, read implementations, explore PRs/MRs
  • Local Tools: Search code (ripgrep), browse directories, find files in your local codebase
  • LSP Intelligence: Go to Definition, Find References, Call Hierarchy — compiler-level understanding
  • Package Discovery: Resolve npm/PyPI packages to their source repos

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

Packages

This is a yarn-workspaces monorepo. Each package has its own README.md; all setup/reference docs live in docs/, and all AI agent guidance lives in the root AGENTS.md.

PackagePurpose
octocode-mcpMCP server — 14 tools across GitHub/GitLab/Bitbucket, local FS, LSP
octocode-cliCLI — installer, tool runner, skills marketplace
octocode-vscodeVS Code extension — GitHub OAuth + multi-editor MCP install
octocode-sharedShared utilities — credentials, session, platform
octocode-security-utilsStandalone security utilities

Skills

Agent Skills are a lightweight, open format for extending AI agent capabilities. Skills index: skills/README.md

Research & Code Analysis

SkillWhat it does
ResearcherCode search & exploration — local LSP + external (GitHub, npm/PyPI)
ResearchMulti-phase research with sessions, checkpoints, state persistence
EngineerUnderstand, write, analyze, audit code — AST + LSP + dependency graph
BrainstormingIdea validation grounded in evidence — GitHub, npm/PyPI, web in parallel
NewsWhat's new in AI, dev tools, web platform, security, notable repos

Planning & Writing

SkillWhat it does
PlanEvidence-based planning: Understand > Research > Plan > Implement
RFC GeneratorFormal technical decisions with alternatives, trade-offs, and recommendations
Doc Writer6-phase pipeline producing 16+ validated docs
Prompt OptimizerTurn weak prompts into enforceable agent protocols
Agentic FlowThinking framework for designing/reviewing MCP & multi-agent workflows

Review & Critique

SkillWhat it does
PR ReviewerPR & local code review across 7 domains with LSP flow tracing
RoastBrutal code critique with file:line citations and severity levels

Build & Output

SkillWhat it does
SlidesPolished multi-file HTML presentations via 6-phase design flow
DesignDynamic DESIGN.md generator covering visual language, components, a11y
Chrome DevToolsCDP-level browser debugging: network, console, perf, DOM, screenshots

Tooling & Setup

SkillWhat it does
InstallInteractive step-by-step Octocode installer for macOS and Windows
CLIRun Octocode MCP tools from the terminal without wiring MCP
Search SkillFind, evaluate, install, refactor Agent Skills (SKILL.md format)
StatsLocal HTML dashboard from Octocode MCP usage stats

https://github.com/user-attachments/assets/5b630763-2dee-4c2d-b5c1-6335396723ec

Documentation

Full index: docs/README.md. All monorepo documentation lives in docs/ — no per-package docs/.

Docs map

  • docs/configuration/ — install, auth providers, MCP clients, env/config, troubleshooting
  • docs/dev/ — tool/API references, workflows, architecture, contributing, skills
  • docs/specs/ — design specs and RFCs

Setup

Tool References

CLI & Skills

Shared Internals

Operations

The Manifest

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

Contributing

See the Development Guide for monorepo setup, testing, and contribution guidelines.

Built with care for the AI Engineering Community
mcpai-assistantgithubgitlabbitbucketcode-searchlspdeveloper-toolscliskills