By shopwareLabs
GitHub CLI MCP servers wrapping gh for pull requests, issues, CI runs, jobs, commits, search, labels, projects, and reviews. Read server (always active) with 30 tools and write server (opt-in) with 23 tools. Includes SessionStart hook that injects MCP tool directives and PreToolUse hooks that enforce MCP tool usage. Configuration-optional: works without config when gh is authenticated.
Experimental Community Project: Maintained by Shopware Labs, not an official Shopware product. Not affiliated with, endorsed by, or sponsored by Anthropic, OpenAI, or any other AI provider. Provider and product names are used only to describe compatibility. Provided as-is without warranty.
GitHub CLI tools for AI coding agents. Wraps the GitHub CLI (gh) behind MCP servers — pull requests, issues, CI runs, jobs, commits, search, labels, projects, and repository file browsing — as first-class MCP tools, with hook-based enforcement that keeps the agent on the tools instead of raw gh bash calls.
The MCP servers use the assistant-neutral Model Context Protocol and can be integrated with MCP-capable coding agents. This repository provides tested github-mcp plugin packaging for both Claude Code and Codex.
Origin: Extracted from its sibling project shopwareLabs/ai-coding-tools — a broader AI-coding-tools marketplace — into this standalone repository. The GitHub tooling was split out so it can evolve and be installed independently.
Packaged plugin requirements: gh CLI authenticated (gh auth login), jq, and either Claude Code or Codex.
/plugin marketplace add shopwareLabs/github-agent-tools
/plugin install github-mcp@github-agent-tools
Restart Claude Code after installation for the MCP servers to initialize.
codex plugin marketplace add shopwareLabs/github-agent-tools
codex plugin add github-mcp@github-agent-tools
Start a new Codex task after installation. Open /hooks to review and trust the bundled enforcement hooks; Codex skips untrusted plugin hooks.
The read server (gh-tooling) is always active. The write server (gh-tooling-write) is opt-in via enable_write_server: true in a .mcp-gh-tooling.json config file. Configuration is optional — the read server works out of the box when gh is authenticated.
Optional Claude Code setup: the companion plugin-setup plugin remains Claude Code-only because it uses Claude Code interaction and permission settings. Install it with /plugin install plugin-setup@github-agent-tools, ask Claude to "set up github-mcp", then uninstall it when setup is complete. Codex users configure .mcp-gh-tooling.json manually as described in the plugin guide.
| Component | Description |
|---|---|
| 🔌 MCP | Two servers — gh-tooling (30 read tools) and gh-tooling-write (23 write tools, gated) |
| 🪝 Hooks | SessionStart directive + PreToolUse enforcement that redirects gh bash calls to the MCP tools |
See plugins/github-mcp/README.md for full configuration, the complete tool reference, and troubleshooting. See plugins/github-mcp/REFERENCE.md for per-tool parameter docs.
# One-time: install BATS locally
./.github/scripts/setup-bats.sh
# Run the suites
.bats/bats-core/bin/bats -r plugin-tests/
MIT.
Admin access level
Server config contains admin-level keywords
Executes bash commands
Hook triggers when Bash tool is used
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub shopwarelabs/github-agent-tools --plugin github-mcpWriting skills for Shopware core contributors: Architecture Decision Records, PR descriptions (trunk and feature-branch formats), and RELEASE_INFO/UPGRADE entries. Analyzes branch diffs, classifies changes, and writes content calibrated to change magnitude.
Knowledge skill for interpreting CI failure logs from Shopware GitHub Actions workflows. Covers PHPUnit, PHPStan, ECS, ESLint, TypeScript, Stylelint, Prettier, Jest, Playwright, ludtwig, and Lighthouse.
Three MCP servers for PHP and JavaScript operations plus Shopware LSP. PHP (php-tooling): PHPStan, ECS, PHPUnit, PHPUnit coverage gap analysis, Symfony Console. Administration (js-admin-tooling): ESLint, Stylelint, Prettier, Jest, TypeScript, Vite builds. Storefront (js-storefront-tooling): ESLint, Stylelint, Jest, Webpack builds. LSP: Service ID completion, Twig templates, snippets, routes, feature flags. Includes SessionStart hook that injects MCP tool directives and PreToolUse hooks that enforce MCP tool usage by blocking bash commands. Supports native, Docker, Docker Compose, Vagrant and DDEV environments.
Semantic code research using ChunkHound's multi-hop search and LLM synthesis
Generate and validate PHPUnit unit tests for Shopware 6. Analyzes source classes to detect category (DTO, Service, Flow/Event, DAL, Exception) and applies category-specific templates. Wave-based Agent Teams orchestration: spawn-per-wave agents with peer-to-peer debate, adversarial red team, and defense rounds. MCP-driven test rules for Shopware compliance. Orchestrator runs inline fix loop (max 4 iterations) with oscillation detection. Bundles MCP server config (requires dev-tooling plugin installed). Configure via .mcp-php-tooling.json.
Harness-native ECC plugin for engineering teams - 67 agents, 279 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.
Persistent coding companion for Claude Code, Pi, and Oh My Pi. MCP-based terminal pet with ASCII art, stats, reactions, and personality.
Semantic search for Claude Code conversations. Remember past discussions, decisions, and patterns.
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.