Turn your AI agents and automations into installable plugins and packages — 12 skills, 10 commands across 3 phases (ASSESS → PACKAGE → SHIP). Covers every major framework (LangChain, CrewAI, AutoGen, Pydantic AI, LlamaIndex, n8n, Flowise, Dify, ComfyUI, browser agents) and every distribution target (MCP servers…
★not rated 3 1mo agoA
tokens not measured
originalApache-2.0
Query Xcode's pre-built symbol index from Claude Code. Surgical, semantic Swift/ObjC symbol lookups — find references, definitions, overrides, conformances, and blast radius without reading full files.
★not rated 3 2mo agoA
tokens not measured
originalMIT
Drive your real, logged-in Chrome from Claude Code, and list tabs across readable Chrome and Chromium profiles. Exposes the extension's browser tools through the Claude in Chrome bridge and adds a read-only listopentabs tool backed by local session data. Driving requires the connected extension and a running Chrome.
★not rated 3 25d agoA
tokens not measured
originalMIT
Skills that keep a codebase healthy as agents work in it: repo orientation, docstring discipline, doc freshness, test coverage for new code, and pre-commit verification gates.
★not rated 3 4mo agoA
tokens not measured
originalMIT
Faithful web research → a folder of clean markdown (aggregation, not synthesis). Claude browses, judges each page, and keeps only the good ones. First run silently installs a headless browser (200MB) under /.local/share/ultra-research.
★not rated 3 1mo agoA
tokens not measured
originalMIT
Two code-review strategies sharing one contract: /consensus-review (parallel multi-source panel with an Opus arbiter, codex consultant and strict sceptic) and /loop-review (sequential coverage-directed passes with evidence-backed cells and a CLEAN verdict enforced in code). Strictly read-only.
★not rated 3
changed 9d agoA
tokens not measured
originalMIT
Cross-tool context bridge between Claude Code and Codex. Logs your conversation and exposes it to the other tool so you can hand off plans, share context, and pick up where the other tool left off.
★not rated 3 26d agoA
tokens not measured
originalMIT
MCP version of the Android debug skill. Exposes adb / emulator / logcat / screenshot / UI-tree as MCP tools; the host LLM (Claude Code) selects each UI action itself. No separate LLM API key required.
★not rated 3 2mo agoA
tokens not measured
originalMIT
Every prompt takes a sip 💧 — a fun Claude Code status line showing how much water talking to Claude costs, plus cost, 5h/weekly usage (mirrors /usage), context, model & effort, and turn count. All local, all free.
★not rated 3 2mo agoA
tokens not measured
originalMIT
Your agent writes the code, a different model tries to break it, and only what survives ships. Cross-model, edit-locked review with a machine-readable verdict, gated merges, durable decisions, stuck-loop and timeout guards, and read-only reasoning one-shots. Harness-neutral.
★not rated 3▲
+1
changed todayA
tokens not measured
originalMIT
Native Claude Code AI code review + blocking pre-push gate on your Claude subscription. Isolated headless reviewer, severity/confidence verdicts, fails closed. open-code-review, done compliantly.
★not rated 3
changed 2d agoA
tokens not measured
originalApache-2.0