Open-source tools for AI-augmented engineering — Claude Code plugins, MCP servers, security scanners, schedulers, and dev-productivity utilities. Each project is its own repo with its own LICENSE, README, tests, and release tags.
awesome-agents(kyrolabs) — PR #483awesome-claude-code-toolkit(rohitg00) — PR #412Awesome-AI-Agents(Jenqyang) — PR #241- Auto-discovered by
claudemarketplaces.com,claude-plugins.dev, andChat2AnyLLM/awesome-claude-plugins - Pending manual submissions: anthropics/claude-plugins-official and mcpservers.org/submit (web forms)
Ten small, focused plugins solving real day-to-day developer-productivity problems. Python 3 stdlib only at runtime; hermetic tests.
| Plugin | What it does | Repo |
|---|---|---|
argus-audit |
White-hat security audit + pentest orchestrator. Scope-gated authz vault, intel layer (NVD/KEV/EPSS), 5 scanners (static / secrets / supply-chain / TLS / HTTP). 174 tests. | mturac/pluginpool-argus-audit |
commit-narrator |
Conventional commit message generator from staged git diff. | mturac/pluginpool-commit-narrator |
pr-storyteller |
PR title + body + test plan from commits/diff vs base branch. | mturac/pluginpool-pr-storyteller |
test-gap |
Surface changed git diff lines that aren't covered by tests. Cobertura / LCOV / coverage.py. |
mturac/pluginpool-test-gap |
deps-doctor |
Dependency vuln/outdated/license audit across npm / pip / cargo / go. Monorepo-aware. | mturac/pluginpool-deps-doctor |
standup-gen |
Daily standup notes from git activity across one or many repos. | mturac/pluginpool-standup-gen |
todo-harvest |
TODO/FIXME/HACK scan with git blame author + age in days. Batched per-file blame. | mturac/pluginpool-todo-harvest |
env-lint |
.env vs .env.example key parity. Inline-comment safe, --strict mode. |
mturac/pluginpool-env-lint |
secret-guard |
Pre-commit secret scanner: regex + entropy, with SHA-redacted output. | mturac/pluginpool-secret-guard |
flaky-detector |
Run a test command N times, report per-test flakiness % with pytest -q imputation. | mturac/pluginpool-flaky-detector |
changelog-forge |
Group conventional commits since last tag into a CHANGELOG section + semver bump suggestion. | mturac/pluginpool-changelog-forge |
| Project | What it does | Repo |
|---|---|---|
| 🪑 claude-roundtable | Multi-agent governance for Claude Code. Council deliberates, votes, dispatches, and enforces quality gates. | mturac/claude-roundtable |
| ⏱️ leyla-scheduler | Durable, session-aware task scheduler in Rust. Jobs survive disconnects. | mturac/leyla-scheduler |
| 🔎 skill-hunter | Pre-execution layer that makes agents check for existing skills before building from scratch. | mturac/skill-hunter |
| 🛡️ promptguard | Audits prompts as behavioral contracts. Pre-write guard for agents that ship code. | mturac/promptguard |
| 🧠 recall-mcp | Shared brain for AI agents. SQLite-backed persistent memory over MCP. Hybrid FTS5 + vector search. | mturac/recall-mcp |
| 👵 moooom-claude | 10 cultures of moms nagging your Claude Code to drink water and sit up straight. | moooom-claude/moooom-claude |
The pluginpool plugins + argus-audit + moooom-claude + skill-hunter expose a .claude-plugin/plugin.json manifest and can be added in one step via the marketplace:
# Add the marketplace to Claude Code
/plugin marketplace add mturac/claude-plugin-marketplace
# Browse and install plugins
/plugin install argus-audit
/plugin install commit-narrator
/plugin install moooom-claude
# …See mturac/claude-plugin-marketplace for the manifest catalog.
- Tests are not optional. Every project ships a test suite that runs offline with no network.
- Standard library first. Python-stdlib-only where possible. External dependencies require justification.
- Plugins are independent repos. Each can be open-sourced, vendored, or installed individually.
- Reviewed before release. Every v0.1.0 tag landed after at least one round of
MiMo/ gemini code review with the findings explicitly addressed.
All projects are MIT licensed unless their own LICENSE file says otherwise. Each repo carries its own copy.