PluginBench
MCP Server
Active
MIT

io.github.Dave-London/github MCP Server

io.github.Dave-London/github

Structured JSON output for GitHub operations—PRs, issues, actions—with token-efficient, schema-validated responses.

What is the io.github.Dave-London/github MCP server?

The GitHub MCP server wraps GitHub CLI operations (gh) and returns clean, schema-validated JSON instead of raw terminal text. It provides tools for managing pull requests, issues, and GitHub Actions with 80–95% fewer tokens than raw output, enabling reliable agent workflows without brittle string parsing.

This server exposes GitHub operations like listing/creating/updating pull requests and issues, checking action runs, and managing workflows. It's part of the Pare toolkit, which eliminates fragile CLI output parsing by returning typed, structured data—agents get consistent field names regardless of platform or tool version, and use significantly fewer tokens per operation.

How to install io.github.Dave-London/github

Copy-paste configuration for popular MCP clients.

transport: stdio
Config generated by PluginBench — verify against the source before use.
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": [
        "-y",
        "@paretools/github"
      ]
    }
  }
}

Tools & capabilities

Tools this server exposes to the agent.

  • PR operations — Create, list, view, update, merge, and close pull requests with structured output
  • Issue operations — Create, list, view, update, and close issues with schema-validated responses
  • GitHub Actions — List and check status of workflow runs and action executions

Use cases

  • Automatically list and filter pull requests by status, author, or label without parsing terminal output
  • Create or update issues programmatically with structured validation and consistent field names
  • Monitor GitHub Actions workflow runs and retrieve their status in token-efficient JSON format
  • Merge pull requests and manage their lifecycle through reliable, typed API responses
  • Search and retrieve issue/PR details with 80–95% fewer tokens than raw gh CLI output

io.github.Dave-London/github MCP server FAQ

What is the GitHub MCP server?

It's an MCP server that wraps the GitHub CLI (gh) and returns structured, schema-validated JSON for pull requests, issues, and GitHub Actions instead of raw terminal text. This eliminates parsing fragility and reduces token usage by 80–95%.

Is it free?

Yes. Pare is open-source (MIT license) and free to use. You only need a GitHub account and the gh CLI installed with valid authentication.

How do I install it in Cursor or Claude?

Use the quick setup: `npx @paretools/init --client claude-code --preset web` (or your relevant preset), then restart your client. Or manually add the server to your MCP config with `npx -y @paretools/github`.

What authentication is required?

You need the GitHub CLI (gh) installed and authenticated. Run `gh auth login` once to set up your credentials. The server uses your existing gh authentication.

Can I limit which tools are available?

Yes. Use environment variables to filter tools: `PARE_GITHUB_TOOLS=pr:list,issue:create` or the universal `PARE_TOOLS=github:pr:list,github:issue:create`. Set via the `env` key in your MCP config.

How much do I save on tokens?

Structured output typically saves 80–95% on tokens compared to raw gh output. For example, a verbose PR list that takes ~500 tokens in raw format may use only ~50–100 tokens as structured JSON.

README (reference)

Source of truth, from the repository.

<h1><img src="assets/logo.png" alt="" width="80" valign="middle" />&nbsp;&nbsp;Pare</h1>

CI codecov npm Downloads TypeScript License: MIT Node.js >= 20 OpenSSF Scorecard OpenSSF Best Practices

Reliable, structured CLI output for AI agents — no more parsing fragile terminal text.

Pare provides MCP servers that wrap common developer tools (git, npm, docker, test runners, etc.) and return clean, schema-validated JSON instead of raw terminal text. Agents get typed data they can act on directly, without brittle string parsing.

The Problem

Parsing CLI output is fragile. Raw terminal text includes ANSI escape codes, decorative headers, progress bars, locale-specific formatting, and platform differences that break agent workflows in subtle ways. An agent that works fine with git status on macOS may fail on Windows because the output format changed. A test runner's summary line might shift between versions, silently breaking a regex.

Pare eliminates this entire class of errors by returning schema-validated JSON with consistent field names, regardless of platform, tool version, or locale. As a bonus, structured output is significantly smaller — agents use fewer tokens per tool call:

Tool CommandRaw TokensPare TokensReduction
docker build (multi-stage, 11 steps)3732095%
git log --stat (5 commits, verbose)4,99238292%
npm install (487 packages, warnings)2414183%
vitest run (28 tests, all pass)1963980%
cargo build (2 errors, help text)43613868%
pip install (9 packages, progress bars)28810165%
cargo test (12 tests, 2 failures)35119046%
npm audit (4 vulnerabilities)28718536%

Token estimates use ~4 chars/token. The biggest savings appear on verbose commands (builds, installs, tests). For simpler tools like eslint or tsc, the main advantage is reliable structured data — agents can use typed JSON directly rather than parsing strings.

How It Works

Each Pare tool returns two outputs:

  • content — human-readable text, for MCP clients that display it
  • structuredContent — typed, schema-validated JSON, ready for agents to process

This uses MCP's structuredContent and outputSchema features to provide type-safe, validated data that agents can rely on without custom parsing.

Example: git status

Raw git output (~118 tokens):

On branch main
Your branch is ahead of 'origin/main' by 2 commits.
  (use "git push" to publish your local commits)

Changes to be committed:
  (use "git restore --staged <file>..." to unstage)
        modified:   src/index.ts
        new file:   src/utils.ts

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git restore <file>..." to discard changes in working directory)
        modified:   README.md

Untracked files:
  (use "git add <file>..." to include in what will be committed)
        temp.log

Pare structured output (~59 tokens):

{
  "branch": "main",
  "upstream": "origin/main",
  "ahead": 2,
  "staged": [
    { "file": "src/index.ts", "status": "modified" },
    { "file": "src/utils.ts", "status": "added" }
  ],
  "modified": ["README.md"],
  "deleted": [],
  "untracked": ["temp.log"],
  "conflicts": [],
  "clean": false
}

50% fewer tokens. Zero information lost. Fully typed. Savings scale with output verbosity — test runners and build logs see 80–92% reduction.

Available Servers (28 packages, 240 tools)

Install only the servers relevant to your stack — most projects need just 2–4. The full catalog covers a wide range of ecosystems so Pare works wherever you do.

CategoryServersToolsWraps
Version Controlgit, github55git, gh
Languages & Packagesnpm, python, cargo, go, deno, bun, nix, dotnet, ruby, swift, jvm101npm, pip, cargo, go, deno, bun, nix, dotnet, gem, swift, gradle, maven
Build, Lint & Testbuild, lint, test, cmake, bazel23tsc, esbuild, vite, webpack, eslint, prettier, biome, vitest, pytest, jest
Infrastructuredocker, k8s, infra, security, remote40docker, kubectl, helm, terraform, ansible, trivy, ssh
Utilitiessearch, http, make, process, db21ripgrep, fd, curl, make, just, psql, mysql, redis, mongosh

Tool Schemas — detailed response examples and field descriptions for every tool. See also Tool Response Examples for quick JSON samples.

Quick Setup

# 1. Configure MCP servers (non-interactive)
npx @paretools/init --client claude-code --preset web

# 2. Add agent rules to your project
#    (append to existing CLAUDE.md, or copy if new)
cat node_modules/@paretools/init/rules/CLAUDE.md >> CLAUDE.md

# 3. Restart your client session

# 4. Validate
npx @paretools/init doctor

Available presets: web, python, rust, go, jvm, dotnet, ruby, swift, mobile, devops, full

Setup Guides by Client

Claude CodeClaude DesktopCursor
VS Code / CopilotWindsurfCline / Roo Code
OpenAI CodexGemini CLIZed
Continue.dev

Full Quickstart Guide — presets, ecosystem mapping, validation

Manual Configuration — config paths and formats for all clients

Agent Integration Guide — rule files, hooks, CLI-to-MCP mapping

Configuration

Tool Selection

By default, every Pare server registers all of its tools. If a server exposes tools you don't need — or you want to limit which tools are available to an agent — you can filter them with environment variables.

Per-server filter — restrict a single server's tools:

# Only register status and log in the git server
PARE_GIT_TOOLS=status,log npx @paretools/git

Universal filter — restrict tools across all servers:

# Only register these specific tools across any server
PARE_TOOLS=git:status,git:log,npm:install npx @paretools/git

Disable all tools — set the env var to an empty string:

PARE_GIT_TOOLS= npx @paretools/git   # no tools registered
Env VarScopeFormatExample
PARE_TOOLSAll serversserver:tool,...git:status,npm:install
PARE_{SERVER}_TOOLSOne servertool,...status,log,diff

Rules:

  • No env var = all tools enabled (default)
  • PARE_TOOLS (universal) takes precedence over per-server vars
  • Server names use uppercase with hyphens replaced by underscores (e.g., PARE_MY_SERVER_TOOLS)
  • Whitespace around commas is ignored

Common patterns:

# Read-only git (no push, commit, add, checkout)
PARE_GIT_TOOLS=status,log,diff,branch,show

# Minimal npm
PARE_NPM_TOOLS=install,test,run

# Only specific tools across all servers
PARE_TOOLS=git:status,git:diff,npm:install,test:run

In JSON MCP config, set via the env key:

{
  "mcpServers": {
    "pare-git": {
      "command": "npx",
      "args": ["-y", "@paretools/git"],
      "env": {
        "PARE_GIT_TOOLS": "status,log,diff,show"
      }
    }
  }
}

Troubleshooting

IssueSolution
npx not found / ENOENT on WindowsUse cmd /c npx wrapper (see your client's setup guide)
Slow first startRun npx -y @paretools/git once to cache, or install globally: npm i -g @paretools/git
Node.js version errorPare requires Node.js >= 20
NVM/fnm PATH issuesUse absolute path to npx: e.g., ~/.nvm/versions/node/v22/bin/npx
MCP connection timeoutSet MCP_TIMEOUT=30000 for Claude Code, or increase initTimeout in client config
Too many tools filling contextUse tool selection env vars to limit tools, or only install the servers you need

Contributing

Each server is a self-contained package. See CONTRIBUTING.md for the full guide.

License

MIT

Related MCP servers

Structured Go tool output (build, test, vet) as typed JSON diagnostics for AI agents.

130
TypeScript
MIT
View repository →

MCP server for HTTP requests with structured, token-efficient JSON output instead of raw curl responses

130
TypeScript
MIT
View repository →

Structured linting output as typed JSON diagnostics for ESLint and format checkers.

130
TypeScript
MIT
View repository →

Structured JSON output for Make/Just task runners—eliminates parsing fragile CLI text for AI agents.

130
TypeScript
MIT
View repository →

Structured npm/pnpm operations (install, audit, outdated, list) as typed JSON for AI agents.

130
TypeScript
MIT
View repository →

Structured build output (tsc, esbuild, vite, webpack) as typed JSON diagnostics for AI agents.

130
TypeScript
MIT
View repository →