All categories
AI & Agents
LLMs, agent workflows, RAG, MCP, prompting, and AI app patterns
- 300,204
- 12,509
Security grades appear on each card once the skill has been scanned. Newly imported skills may briefly show without a grade until the backfill job runs.
Open in full browserBrowse ai & agents skills
Showing 337–360 of 300,204 skills
- Aiui McpBefore writing a yes/no question, a numbered option list, or a multi-question request into the chat, open a native desktop dialog instead — `confirm` for yes/no (always for delete/force-push/drop/deploy), `ask` for one-of-N with per-option context, `form` for ≥ 2 related inputs / secrets / dates / sliders / sortable lists / table-row triage / image confirm, `compare` for picking one of 2–3 full variants side by side, `gallery` for a per-item verdict on a batch of images/videos, `notify` for a...Votes: 0GitHub stars: 6
- DocsRender native desktop dialogs on the user's machine via aiui's MCP server — `confirm` before destructive actions (delete, drop, force-push, deploy), `ask` for pick-one-of-N where context per option matters, `form` for multi-input requests, secrets, dates, sliders, sortable lists, or image confirmation, `compare` for A/B(/C) side-by-side picks, `gallery` for batch image/video review, `notify` for a fire-and-forget completion signal that doesn't block on a reply.Votes: 0GitHub stars: 6
- Testing With M3Use when setting up M3 or when writing, running, debugging, or evaluating tests for an MCP server or for agents that use one. Covers m3 init, setup, doctor, and test; direct MCP tests; agent tests with Codex, Claude Code, OpenCode, Pi, or ACP agents; evaluations and LLM judges; saved runs, feedback files, and traces. Use it whenever a project has an m3.toml file, imports m3, or the user mentions M3, MCP server tests, or testing an MCP tool with an agent.Votes: 0GitHub stars: 7
- AgentsHow m3 init and m3 setup install the testing-with-m3 agent skill, and how to install it yourself.Votes: 0GitHub stars: 7
- Macos HarnessSee and operate macOS apps, iOS Simulators and Android emulators and phones from the shell with the macos-harness CLI - list apps and windows, read an app's UI as refs, take screenshots, press buttons, type, choose menu items, drag, scroll, swipe and right-click; boot simulators and emulators, build and run apps on them. Use when a task needs a Mac, iOS or Android app's UI, such as checking an app you built or driving one of Apple's apps.Votes: 0GitHub stars: 4
- Veto Policy RuntimeCreate and apply new Veto policies safely for AI agents using deterministic rules first, with optional LLM-assisted generation when needed. Use this when you must add guardrails without editing or deleting existing policies.Votes: 0GitHub stars: 14
- ExpectUse when editing .tsx/.jsx/.css/.html, React components, pages, routes, forms, styles, or layouts. Also when asked to test, verify, validate, QA, find bugs, check for issues, or fix expect-cli failures.Votes: 0GitHub stars: 14
- Debug AgentSystematic evidence-based debugging using runtime logs. Generates hypotheses, instruments code with NDJSON logs, guides reproduction, analyzes log evidence, and iterates until root cause is proven with cited log lines. Use when the user reports a bug, unexpected behavior, or asks to debug an issue.Votes: 0GitHub stars: 14
- ResumeValidate a saved native IdeaScientist run and continue its next admissible stage without resetting budgets or overwriting outputs.Votes: 0GitHub stars: 3
- IdeateStart a bounded native IdeaScientist run for a research problem, grounded gaps, one intuition and an independently reviewed proposal.Votes: 0GitHub stars: 3
- EvaluateIndependently evaluate a native proposal with explained quality scores, coverage-limited novelty and a previously frozen blind baseline.Votes: 0GitHub stars: 3
- CheckRun the standalone offline validator for a native IdeaScientist run's artifacts, evidence receipts and stage dependencies.Votes: 0GitHub stars: 3
- Ideascientist ResumeContinue a saved IdeaScientist run with its spent budgets and exact dependencies.Votes: 0GitHub stars: 3
- Ideascientist IdeateStart automatic grounded scientific ideation for a research problem.Votes: 0GitHub stars: 3
- Ideascientist EvaluateIndependently evaluate a completed IdeaScientist proposal.Votes: 0GitHub stars: 3
- Ideascientist CheckCheck saved IdeaScientist artifacts using the canonical offline validator.Votes: 0GitHub stars: 3
- Spring Boot ConventionsHouse conventions for writing Spring Boot (Java) backend code, covering controllers, services, DTOs, pagination, error responses. Use whenever the user asks to write, add, refactor, or review a Spring Boot controller, service, repository, DTO, endpoint, or REST API in Java. Trigger phrases - "Spring Boot", "REST controller", "@RestController", "endpoint", "service class", "DTO", "JPA repository". Do NOT use for Vue, Pinia, TypeScript, or frontend work.Votes: 0GitHub stars: 30
- Write CaseAuthor or fix an eval case (prompt.md, graders/*.md, case.yaml) for a Claude Code plugin, skill, or hook in the official `claude plugin eval` format. Use when the user says "write an eval for", "add a test case for this skill/hook", "my grader is wrong", or "how do I assert the hook fired". Do not use to run the suite or explain a red result (that is run), or to fix the setup after a regression (that is repair).Votes: 0GitHub stars: 30
- SetupSet up regression testing (CI) for this repository's Claude Code configuration (plugin, skills, hooks, CLAUDE.md). Use when the user says "set up config-drift-checker", "add evals for my plugin/skills/hooks", "test my Claude Code setup", "make sure my hooks keep working after updates", or "wire the eval GitHub Action". Do NOT use for evaluating an LLM application or prompts; this is for agent configuration only.Votes: 0GitHub stars: 30
- RunRun the agent-config eval suite locally and compare with the stored baseline; explain why a case went red. Use when the user says "run the evals", "run config-drift-checker", "did the Claude Code update break my setup", "why is the eval red", or "compare against baseline". Do not use to change the setup itself (that is repair) or to write new cases (that is write-case).Votes: 0GitHub stars: 30
- RepairAfter a red eval run, propose the smallest change to the agent setup (CLAUDE.md, a skill, a hook) that restores the drifted behaviour, verify it by re-running only the failing cases, and write a PR-ready summary. Use when the user says \"fix the regression\", \"repair the setup\", \"make the eval green again\", or when the config-drift-checker Action runs with repair: true. Do not use to run the suite or diagnose a red result (that is run), and never to edit eval cases or graders.Votes: 0GitHub stars: 30
- Agentvm DevBuild, run and test agentvm fast, try a change on real VMs in an isolated sandbox server, and check what is inside a VM. Use when working on agentvm's server, dashboard, guest scripts or VM helper and you need to build, test, run or verify something.Votes: 0GitHub stars: 2
- Worktree JanitorSweep .claude/worktrees/ across every repo in the workspace — list each worktree's branch, dirtiness, age, and merge state; remove the safe ones, flag the rest. Use when the user says "/worktree-janitor", "clean up worktrees", or a checkout fails with "already used by worktree".Votes: 0GitHub stars: 3
- QueueQueue a prompt to run after the current response finishes. Usage - /queue <prompt>, /queue list, /queue clear. Handled by hooks; user-invoked only.Votes: 0GitHub stars: 3