Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Initialize or repair SigNoz MCP server configuration for Agent Plugins v1, Claude Code, Codex, Cursor, VS Code/GitHub Copilot, Claude Desktop, Gemini CLI, Devin CLI, Grok Build, Windsurf, Zed, Antigravity CLI, OpenCode, or another MCP client. Use this skill before any SigNoz docs, query, dashboard, alert, or view workflow when `signoz_*` tools are unavailable, or when the user says "setup SigNoz MCP", "configure SigNoz plugin", "wrong region", "change SigNoz region", "MCP auth failed", or asks to connect SigNoz Cloud or a self-hosted MCP endpoint, even if they do not mention the plugin.
This skill should be used for the tool-interface layer of an agent system specifically: writing tool descriptions agents can route on, designing tool schemas and response formats, naming conventions, actionable error recovery messages, MCP server design, tool-set consolidation, and deciding when to add or remove an individual tool. Use this when the unit of work is a single tool or a set of tools. Route project-shape, pipeline architecture, and task-model-fit decisions to project-development; route deciding whether to introduce sub-agents to multi-agent-patterns.
BLOCKING REQUIREMENT — When a user requests adding, building, creating, or implementing ANY feature or component for a Wix CLI app, you MUST invoke this skill IMMEDIATELY as your absolute FIRST action — when exploring or reading files, BEFORE launching any agents — make sure this skill is loaded. Trigger on add, build, create, I want, implement, help me, where X is any feature/component. Non-negotiable — invoke immediately upon recognizing a Wix feature build request.
Run AI apps via inference.sh CLI - image generation, video creation, LLMs, search, 3D, Twitter automation. Models: FLUX, Veo, Gemini, Grok, Claude, Seedance, OmniHuman, Tavily, Exa, OpenRouter, and many more. Use when running AI apps, generating images/videos, calling LLMs, web search, or automating Twitter. Triggers: inference.sh, infsh, ai model, run ai, serverless ai, ai api, flux, veo, claude api, image generation, video generation, openrouter, tavily, exa search, twitter api, grok
Use when the user asks to draft, scope, or write up a new Strikethroo plan for a work order, feature, or initiative in this repository. Triggers include create plan, new plan, plan this, scope a plan, and strikethroo plan. Do not use to decompose an existing plan into tasks, execute a plan, or for generic brainstorming outside Strikethroo.
Use when the user asks to run, execute, or implement one specific task ID within a Strikethroo plan in this repository. Triggers include execute task, run task, implement task, and do task N. Do not use to execute a whole plan or blueprint, generate tasks, or for generic development outside Strikethroo.
Use when the user asks to decompose, break down, or generate tasks for an existing Strikethroo plan ID in this repository. Triggers include generate tasks, break down the plan, decompose plan, and create the task blueprint. Do not use to create a new plan, execute tasks, or for generic project planning outside Strikethroo.
Use when the user asks to review, refine, improve, interrogate, pressure-test, or update an existing Strikethroo plan by plan ID in this repository — triggers include refine plan, improve plan, review plan, red-team the plan, update plan. Do not use to create a new plan, to generate tasks, or for generic brainstorming outside Strikethroo.
Design and build AI agents for any domain. Use when users: (1) ask to "create an agent", "build an assistant", or "design an AI system" (2) want to understand agent architecture, agentic patterns, or autonomous AI (3) need help with capabilities, subagents, planning, or skill mechanisms (4) ask about Claude Code, Cursor, or similar agent internals (5) want to build agents for business, research, creative, or operational tasks Keywords: agent, assistant, autonomous, workflow, tool use, multi-step, orchestration
Execute long-running, multi-session tasks autonomously using Claude Code headless mode or in-session hook-based loops. Supports structured task decomposition (for complex projects) and lightweight Ralph-style iteration (for TDD, bug fixing, refactoring). Use this skill whenever the user says "autonomous", "long-running task", "multi-session", "run this in the background", "keep working on this", "batch process", "iterate until done", "ralph loop", or wants any task that requires sustained, unattended execution.
Bootstrap and operate an LLM Wiki: a persistent Markdown knowledge base with immutable raw sources, an LLM-maintained wiki, a SCHEMA.md operating contract, thin runtime pointer files, EXTEND.md preferences, and optional local BM25 search. Use when the user asks to create or bootstrap an LLM wiki, ingest sources, query or lint a wiki, configure EXTEND.md preferences, or set up BM25/full-text search for an LLM Wiki.
Show real token usage and estimated savings for the current session. Reads directly from the Claude Code session log — no AI estimation. Triggers on /pordee-stats. Output is injected by the mode-tracker hook; the model itself does not compute the numbers
Use when the user asks to run, execute, implement, or carry out a Strikethroo plan or its blueprint by plan ID in this repository — triggers include execute blueprint, run the plan, implement plan, build the plan. Do not use to create a plan, generate tasks only, run a single task, or for generic development outside Strikethroo
When a solution seems impossible and the path forward is blocked. Refuses to take no for an answer: classifies what kind of "no" you hit, then runs lateral-thinking techniques (assumption autopsy, inversion, first principles, constraint toggling, analogical transfer, and more) until the wall cracks or is proven load-bearing. Generates at least 10 angles before evaluating any. AGENTS: use this on yourself when an API doesn't support what you need, an approach has failed twice, or a "that's not possible" sentence is forming — run the fast path BEFORE reporting a dead end. Archives walls to ~/.config/makerskills/unstuck/archive/. Sits upstream of decide. Triggers on "/unstuck," "I'm stuck," "this seems impossible," "we hit a wall," "there's no way to," "they said no," "dead end," "out of options," "I've tried everything," "work around this."
Use the local Codex CLI as an independent second agent. Two branches: (1) proactively run `codex review` for a second opinion after completing a substantive change, before presenting it as done or committing; (2) delegate a well-defined implementation task via `codex exec` only when the user explicitly asks for Codex to do it. Also covers how to prompt Codex.
Use when the user asks to run the complete end-to-end Strikethroo workflow for a work order in one shot in this repository—triggers include full workflow, end-to-end, plan and execute, do everything, or run the whole strikethroo workflow. Do not use when the user wants only one stage (create a plan, generate tasks, or execute a blueprint); use the dedicated skill for that stage instead.
Break large features into independently verifiable, human-reviewable slices under a feature directory in specs/. Use for risky or multi-step feature work that needs upfront questioning, API seams, browser-playable checkpoints, HTML visualizations, screenshot gates, staged implementation plans, recursive fog-of-war reslicing, or proactive research into reference implementations/best practices before slicing. Pairs with your project's verification harness and screenshot gates (the browser checkpoints), [refactor-clean](../refactor-clean/SKILL.md) (review the materialized spec so the plan describes one-owner architecture, not the feature bolted on), [screenshot-critique](../../visual/screenshot-critique/SKILL.md) and [compare-screenshots](../../visual/compare-screenshots/SKILL.md) (the visual gates), and a code-review pass (audit each slice before it lands).
Persistent, portable memory for AI builders and agent teams. Use BEFORE any non-trivial work to retrieve prior decisions, patterns, and gotchas; use AFTER finishing to record what was learned. ALSO use to authenticate with ByteRover — when the user says to authenticate / auth / log in / sign in / connect / link ByteRover (or log out / sign out / disconnect), run the auth flow. Iron Law: query before you think, curate after you implement
This skill should be used when creating a Claude Code slash command. Use when users ask to "create a command", "make a slash command", "add a command", or want to document a workflow as a reusable command. Essential for creating optimized, agent-executable slash commands with proper structure and best practices.
Improve (or bootstrap) an AGENTS.md / CLAUDE.md file using `<important if>` conditional blocks so the agent actually attends to the right guidance at the right time. Use this skill whenever the user mentions AGENTS.md, CLAUDE.md, agent instructions, project rules for AI, "my claude config", onboarding docs for agents, or asks to tighten / shorten / audit / rewrite an existing one — even if they don't explicitly say the filename. Also use when the user complains that an agent keeps ignoring their project rules.
Builds and maintains an LLM-curated knowledge base from papers, articles, transcripts, notes, and project findings. Ingests sources into linked Markdown, answers questions with citations, lints or upgrades a wiki, and preserves useful synthesis across sessions and agents. Also captures verified task experience, consolidates success and failure patterns, and proposes, evaluates, applies, or rolls back evidence-linked skill improvements. Trigger on “add this to my wiki,” “what does the wiki say,” “learn from this task,” “improve this procedure,” /wiki:learn, or /wiki:evolve. Works on a local folder or a hosted wiki URL (MCP). Uses index-first navigation, atomic pages, and hybrid search.
Multi-agent autonomous startup system for Claude Code. Triggers on “Loki Mode.” Orchestrates 100+ specialized agents across engineering, QA, DevOps, security, data/ML, business operations, marketing, HR, and customer success. Takes a PRD to a fully deployed, revenue-generating product with zero human intervention. Features a Task tool for subagent dispatch, parallel code review by 3 specialized reviewers, severity-based issue triage, a distributed task queue with dead-letter handling, automatic deployment to cloud providers, A/B testing, customer feedback loops, incident response, circuit breakers, and self-healing. Handles rate limits with distributed state checkpoints and automatic resume with exponential backoff. Requires the --dangerously-skip-permissions flag
Catch-all for anything CawCut-App related via the local `cawcut` CLI: running published Apps, and open-ended requests like "what can <app> do", "what does <app> generate", "what apps are there", "list official apps", browsing/listing available Apps, or checking an App's inputs/outputs. Use when: "/cawcut-app-run", "run my CawCut app", "invoke published app", "list available CawCut apps", "what apps are available", or the user describes a workflow/template-style creative request (ads, campaigns, multi-step templates) that may match a published app. Fetches the live App catalog and scores intent before routing; hands off simple one-step text-to-image/video or image-to-image/video generation with no App match to cawcut-generate, which will hand back here if it later finds a matching App. Requires `cawcut` installed and authenticated. NOT for: OAuth consent alone, or Apps the user cannot access. When unsure whether a request needs an App or raw generation, start here — this skill decides and delegates.
Research, create, improve, migrate, evaluate, package, install-check, govern, and safely publish qiaomu-flavored agent skills from workflows, prompts, transcripts, docs, SOPs, runbooks, scripts, or notes. Use for new or existing skills, prior-art synthesis, routing/trigger boundaries, trigger or output evals, Skill IR, release gates, README/Profile preparation, GitHub repository and pull-request publication, versioned Releases, clean npx installation, team reuse, and create-and-publish flows. The publication path is self-contained and forbids direct default-branch pushes. Exclude one-off summaries, translations, ordinary docs, non-skill package publishing, and tasks that explicitly should not become a skill.
Drive a REAL browser and native apps the way a user would — not fetch raw HTML — through the persistent `browse` CLI: one long-lived headless Chromium (plus iOS, Android, and macOS apps via accessibility APIs) where goto, snapshot, click, fill, JavaScript, and screenshot share session state at ~100ms/command and interaction rides stable @eN refs, not brittle selectors. Use when verifying, navigating, or automating live web or app behavior beyond what code-reading or a raw fetch reveals
Extracts key learnings from conversations, debugging sessions, and failed attempts. Use at session end or after solving complex problems to capture insights. Stores discoveries in memory (via amplihack.memory.discoveries), suggests PATTERNS.md updates, and recommends new agent creation. Ensures knowledge persists across sessions via Kuzu memory backend.
Publish an agent skill to GitHub in one click. Automatically validates SKILL.md, generates or checks a more engaging README, distinguishes the skill name from the GitHub repo name, creates or updates the repository, pushes it, and verifies it by installing it with npx skills. Activates when the user says “发布这个skill,” “publish skill,” “把skill发到GitHub,” “分享这个skill,” or “/publish-skill.” Supports publishing a skill from the current directory or a specified path.
Generate hyper-detailed prompts for Claude Code agent teams based on a plan, feature request, or conversation context. Use this skill whenever the user wants to create a prompt for an agent team, prepare a task for a coding agent, generate a team prompt, build an autonomous agent workflow, or says things like "create a prompt for agent teams," "prepare a task for a fresh session," "generate a team prompt," "build me an agent prompt," "I want to hand this off to an agent team," "create agent team instructions," or "write me an autonomous coding prompt." Also trigger when the user describes a complex feature and asks to delegate it, wants to prepare work for parallel agents, or mentions "fresh context," "copy-paste prompt," or "autonomous implementation."
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.