Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Run a coding-board control plane when the real job is managing bounded coding cards, isolated workspaces/worktrees, review queues, retries, and PR handoff — not just prompting one agent once. Use when the user needs a visual board or tracker-linked workspace loop for parallel coding tasks, agent comparison, or review-heavy handoff. Triggers on: kanban board, coding board, workspace board, review queue, tracker sync for coding tasks, worktree board, parallel coding agents, card-based agent workflow, and vibe-kanban. Route planning-only requests to `task-planning` / `survey` / `plannotator`, browser/UI execution to `agentation` / `browser-harness` / `playwriter`, and non-code PM / marketing / game-production coordination to their domain skills
Migration guide from @openrouter/sdk to @openrouter/agent for callModel, tool(), stop conditions, and agent features. This skill should be used when code imports callModel, tool(), or stop conditions from @openrouter/sdk and needs to migrate to @openrouter/agent.
The only OpenRouter CLI built for fleet operations: budgets, anomaly alarms, and failover maps over a local usage ledger — not another chat wrapper. Trigger phrases: `which model is cheapest for this task`, `what did my crons spend on openrouter`, `check openrouter credits runway`, `is openrouter degraded right now`, `why did my openrouter cost spike`, `use openrouter`, `run openrouter`.
Mid-conversation reflection skill that pauses execution and zooms out from detail mode to honestly reassess direction, assumptions, and bias. Use when the user says “reflect,” “take a step back,” “step back,” “zoom out,” “are we missing something,” “bigger picture,” “sanity check this,” “are we on track,” “are we overthinking this,” “forest for the trees,” or any variation signaling intent to break out of detail mode and reassess. Also trigger when the conversation has gone deep on implementation details without a strategic check-in, or when the user shows signs of being stuck — that’s often a signal the framing needs a reset, not more detail work. Intentionally low-intake: runs the 5-dimension analysis immediately when prior context is rich enough; asks one forcing clarifier only when invocation context is too thin to reassess from
Route reusable skill-improvement work into one bounded repo-local ratcheting packet: ratchet eligibility, benchmark readiness, loop charter freeze, baseline scoring, one-change mutation, support-surface sync, or final keep/revert report. Use when an existing `SKILL.md`, SOP, prompt, or workflow doc may need sharper triggers, clearer instructions, better support files, or cleaner discovery wording and you want a frozen evaluation harness, append-only experiment logs, and explicit keep-or-revert decisions instead of ad hoc rewriting. Also use when you need to prove that no ratchet is justified yet. Not for GPU-bound Karpathy `autoresearch` runs or hosted app-scale eval / observability platforms such as LangSmith, Braintrust, Weave, or Promptfoo.
For contributors to angular-best-practices. Reviews and audits rule files for accuracy, validity, and formatting compliance. Not for Angular application development — install angular-best-practices instead.
Design domain-specific agent teams, define specialized agents, and generate the skills they use. Use when you need to decompose a complex project into coordinated multi-agent teams, choose the right architecture pattern (pipeline, fan-out/fan-in, expert pool, producer-reviewer, supervisor, hierarchical delegation), generate .claude/agents/ and .claude/skills/ files, or validate and iterate on generated harnesses. Triggers on: harness, build a harness, design agent team, agent team architecture, multi-agent skill generation, set up harness, harness engineering, domain agent team, harness for this project.
Use this skill when the user wants to “create a skill,” “add a skill to a plugin,” “write a new skill,” “improve a skill description,” or “organize skill content,” or needs guidance on skill structure, progressive disclosure, or skill development best practices for Claude Code plugins.
Forge new Claude Code skills or optimize existing ones to V2. Two modes — `forge` scaffolds a new skill (frontmatter, progressive disclosure, helper scripts, mandatory Gotchas, iterate-then-extract); `optimize` makes an existing skill measurably better at its OUTCOME, not just its packaging (quality audit + domain outcome-research + changelog + V1-vs-V2 verification). Use when the user wants to create, write, build, scaffold, improve, upgrade, or optimize a skill.
Provides comprehensive CLAUDE.md file management capabilities including auditing, quality assessment, and targeted improvements. Use when user asks to check, audit, update, improve, fix, maintain, or validate CLAUDE.md files. Also triggers for "project memory optimization", "CLAUDE.md quality check", "documentation review", or when CLAUDE.md needs to be created from scratch. This skill scans all CLAUDE.md files, evaluates quality against standardized criteria, outputs detailed quality reports with scores and recommendations, then makes targeted updates with user approval.
Routes any sales, marketing, ad, or GTM objective to the right specialized skill and outputs the install command for that skill plus a ready-to-paste prompt packed with the user's context. Asks clarifying questions when the objective is ambiguous, then hands back a copy-paste-runnable next step. Covers prospecting, outbound cadences, deals, proposals, forecasting, deliverability, enrichment, intent, content, coaching, CRO, SEO, launch directories, newsletters, email/SMS/push marketing, chatbots, influencer marketing, social media, employee advocacy, media relations, reviews, data hygiene, B2B advertising, retargeting, affiliate, loyalty, digital products, memberships, webinars, checkout, and platform-specific help. Use when the user has a sales or marketing question and isn't sure which skill to use, or wants a multi-skill sequence with a batch install command. Do NOT use to solve problems directly—this skill only routes.
Use when creating or refining a structured Java implementation plan from trusted issue summaries, approved designs, ADRs, OpenSpec changes, existing plans, or a valid combination. The plan records its source artifacts and derivation direction and can remain the execution artifact without requiring OpenSpec. This should trigger for requests such as Create a plan from an issue; Create a plan from OpenSpec; Design an implementation plan; Refine an existing plan. Part of Plinth Toolkit
Entry point and governing principle of the Qiushi methodology. Load once at the start of each new top-level session to constrain all judgments with “seek truth from facts,” and route tasks to the corresponding methodology skill only when clearly applicable. Sub-agents delegated a single, specific task should skip it. English: Entry point and governing principle of the Qiushi methodology. Load once at the start of a top-level session to enforce "seek truth from facts" and route to a downstream skill only when it clearly helps. Delegated sub-agents on narrow tasks should skip it.
Create and refine OpenCode agents through guided Q&A. Use proactively for agent creation, performance improvement, or configuration design. Examples: - User: “Create an agent for code reviews” → ask about scope, permissions, tools, and model preferences; generate AGENTS.md frontmatter - User: “My agent ignores context” → analyze description clarity, allowed-tools, and permissions; suggest improvements - User: “Add a database expert agent” → gather requirements, set convex-database-expert in subagent_type, and configure permissions - User: “Make my agent faster” → suggest smaller models, fewer allowed-tools, and tighter permissions
Write, run, and analyze structured test suites for Agentforce agents — functional AND security. TRIGGER when: user writes or modifies test spec YAML (AiEvaluationDefinition); runs sf agent test create, run, run-eval, or results commands; asks about test coverage strategy, metric selection, or custom evaluations; interprets test results or diagnoses test failures; asks about batch testing, regression suites, or CI/CD test integration; requests security testing, OWASP LLM Top 10, red-teaming, penetration testing, prompt-injection tests, a security grade, or a vulnerability assessment of an agent. DO NOT TRIGGER when: user creates, modifies, previews, or debugs .agent files (use agentforce-generate); deploys or publishes agents; writes Agent Script code; uses sf agent preview for development iteration; analyzes production session traces (use agentforce-observe); performs a static safety review of .agent file content (use agentforce-generate Section 15).
Orchestrate multiple worker agents to implement groomed tasks in Gitea repositories. Use when multiple ready tasks need implementation, when you want autonomous multi-task execution, or when coordinating batch development work with Gitea. Keywords: coordinator, orchestrator, multi-task, parallel, workers, batch, autonomous, gitea, tea.
Governs the Claude Code/Codex Skill surface: reconciles source/install/catalog-visibility drift without losing cold capability. Use when users ask how many Skills load, why the catalog is truncated, want routers visible while other bundles stay cold, or need duplicate/superseded-Skill or old-cache cleanup. Not for building a marketplace manifest (use marketplace-dev) or repo audits (use marketplace-health-check).
Generate typed TypeScript SDKs for AI agents to interact with MCP servers. Converts JSON-RPC curl commands to clean function calls. Auto-generates types, client methods, and example scripts from MCP tool definitions. Use when building MCP-enabled applications, need typed programmatic access to MCP tools, or creating reusable agent automation scripts.
Analyze production Agentforce agent behavior using session traces and Data Cloud, and manage Agent Health Monitoring (AHM) alerts. TRIGGER when: user queries STDM session data or Data Cloud trace records; investigates production agent failures, regressions, or performance issues; asks about session traces, conversation logs, or agent metrics; wants to reproduce a reported production issue in preview; runs findSessions or trace analysis queries; creates, lists, updates, or deletes an AHM data alert on an agent metric (escalation rate, deflection, etc.); asks why an alert is not firing or wants to check whether alerts have fired (notification counts, or the per-alert Incidents view). DO NOT TRIGGER when: user creates, modifies, or debugs .agent files during development (use agentforce-generate); writes or runs test specs (use agentforce-test); uses sf agent preview for local development iteration; deploys or publishes agents.
Pre-implementation confidence assessment (≥90% required). Use before starting any implementation to verify readiness with duplicate check, architecture compliance, official docs verification, OSS references, and root cause identification.
Use when the user asks to enhance, improve, refine, rewrite, strengthen, or validate a prompt, or says "make this prompt better". Returns a clearer, more specific, better structured version of the prompt without executing it, picking validation checks from the prompt's own context. Contexts include coding, research, writing, image or video generation, agentic tasks, data work, and system prompts. Flags ambiguity, contradictions, missing constraints, missing success criteria, format gaps, and leaked secrets. Do not use when the user wants the prompt's task performed.
Your pathfinder for navigating unknown codebases. Investigates with precision, implements surgically, and never assumes — if it doesn't know, it says so. Maintains a .notebook/ knowledge base that grows across sessions, turning every discovery into lasting intelligence. Summons available skills, MCPs, and docs when the mission demands. Use when fixing bugs, implementing features, refactoring, investigating flows, or any development task in unfamiliar territory. Triggers on "fix this", "implement this", "how does this work", "investigate this flow", "help me with this code". Do NOT use for greenfield scaffolding, CI/CD, or infrastructure provisioning.
Independently witness that an Allium loop's convergence claim is true and was reached honestly. Use when the user wants to verify a loop's self-report, confirm tests really pass and no generated test was weakened, produce a convergence certificate or witness record, gate CI on a trustworthy signal, or check that an autonomous run did not cheat its way to green.
Use when creating or updating OpenSpec artifacts from an issue, plan, approved design, ADRs, existing OpenSpec, or a valid combination. The workflow assesses reviewable scope, records source authority and derivation, handles conflicts, and prevents silent synchronization. Triggers include Create OpenSpec from an issue; Convert a plan into OpenSpec; Update an OpenSpec change; Split broad requirements into reviewable changes. Part of Plinth Toolkit
Use the Jev AI decision API from an AI coding agent when a workflow needs a typed choice, score, or yes/no judgment for routing, classification, verification, context compression, or a safety gate. On first use, guide the user through language, API key, and help configuration. Do not use it for open-ended writing, deterministic rules, exact lookups, or final permission enforcement.
Build a retrieval-optimized knowledge layer over agent documentation in dotfiles (.claude, .codex, .cursor, .aider). Use when asked to "optimize docs", "improve agent knowledge", "make docs more efficient", or when documentation has accumulated and retrieval feels inefficient. Generates a manifest mapping task-contexts to knowledge chunks, optimizes information density, and creates compiled artifacts for efficient agent consumption.
Automate browsers using the agent-browser CLI. Use for tasks that require browser control, such as checking in, filling out forms, taking screenshots, and extracting information. Triggers: (1) the user requests browser automation; (2) checking in, filling out forms, or clicking buttons is needed; (3) web content must be retrieved as research material.
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Build CLI scripts alongside MCP servers for terminal environments. File I/O, batch processing, caching, richer output formats. Templates for TypeScript scripts and SCRIPTS.md. Use when: MCP companion scripts, batch processing, saving results to files, CLI API wrappers. Troubleshoot: context too large, no file access, batch input.
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.