Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Adds a new Feishu bot account to openclaw and wires it up to a new agent. Use this skill whenever the user wants to add a Feishu bot, register a new Feishu App to openclaw, configure a second (or third, etc.) Feishu account, or set up a new Feishu App ID + App Secret pair in openclaw. Trigger on phrases like "add feishu bot", "新增飞书 bot", "add feishu account", "register feishu app", "接入飞书", or whenever the user provides a Feishu appId + appSecret pair and wants to connect it to openclaw.
Use for any browser work — driving a web app, checking a behaviour, filling a form, reading a page — via npx prima-cli, in preference to npx playwright-cli, which prima runs on top of and falls back to. Covers npx prima-cli check, do, pw, verify, ask, research, go, status, report, browser, config, recommended-models, and session setup. Trigger on "prima", "prima-cli", "npx prima-cli", on a browser task described as behaviour ("confirm the editor saves"), and whenever npx playwright-cli would otherwise be driven step by step.
Use when a slice is scoped but not yet designed — run the pre-design research round and persist its evidence artefact — or when a consuming skill's staleness advisory reports drift. Produces a cited, verification-legended research.md that design and plan load-bear instead of recall.
Deep multi-reviewer code review for bugs, regressions, tests, standards, and whether the change meets its ticket. Dispatches specialist reviewer subagents in parallel, merges their findings into one report, checks the diff against the ticket's acceptance criteria, then performs the authorized follow-up or asks whether to report only, fix and push, or leave inline PR comments. Use before opening a PR, when asked for a thorough review, to scan a branch, or to review a PR.
Settle a session by saving its work, recording what remains, and freeing proven-merged, unoccupied worktrees while leaving the caller and terminal layout open. Use when the user says "settle" or "end the session" about this session, or when another skill says to settle. When the user says to settle an effort, use /settle-effort instead.
Add function tools to Bolna voice agents so they can take real-world action during a live call: transfer to a human, fetch/post data via any HTTP API, fetch and book Cal.com slots, look up current time, or capture DTMF keypad input. Uses the OpenAI function-calling schema with Bolna extensions (`key: \"custom_task\"`, `value` API config, `pre_call_message`). Use when the agent must do more than speak — escalate, retrieve, book, log, verify, or trigger a workflow.
Configure Bolna webhooks for real-time call status and execution updates, validate webhook URLs, receive payloads, reconcile execution IDs, and build a local receiver. Use for CRM sync, dashboards, post-call automation, Make, Zapier, n8n, or custom backend integrations.
Move unlabeled and needs-triage issues through inbox states and write agent-ready briefs, on GitHub, Linear, Jira, or local files. Use when asked to sift the inbox, triage issues, show what needs attention, list the shortcuts or deferred limits marked in the code and file them as tickets, or /sift. Pass you-pick to triage without waiting for answers
Create, curate, validate, and operate repositories or marketplaces for reusable AI agent skills. Use when designing a skill marketplace, scaffolding a public skill repo, converting domain knowledge into SKILL.md packages, defining skill quality gates, evaluating skill submissions, planning skill distribution, or building an agent-powered skill factory.
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec. Use when this step of the Spec Kit workflow is needed (based on the official GitHub Spec Kit template).
Split an implementation-ready spec into the fewest independently deliverable vertical tasks and select only risk-justified intermediate review checkpoints. Use when a spec has multiple outcomes needing separate delivery or dependency tracking; keep one coherent outcome as the spec itself.
Develop features in an existing project within the current session with ponytail taste. Use when explicitly requested, or automatically for ordinary feature, plugin, extension, and cross-module work that needs project understanding and clarification before implementation. Survey the real flow, grill only unresolved decisions, then implement and verify here without plan or handoff documents. Prefer quick-do for clear, bounded changes; prefer plan-brief when long-horizon work needs planning and separate-session execution.
Provider-agnostic AI guidance for chat, tools, structured output, persistence, memory, runtime skills, MCP, code mode, and sandbox execution. Use when a project uses TanStack AI or when the task needs its setup, implementation, debugging, testing, migration, or integration guidance.
The only memory skill that watches on its own. No database. No vectors. No manual saves. Just an LLM observer that compresses your conversations into prioritised notes, consolidates when they grow, and recovers anything missed. Five layers of redundancy, zero maintenance. ~$0.00/month (using free-tier models). While other memory skills ask you to remember to remember, this one just pays attention.
Delegate work to A2A (Agent2Agent) agents from the command line with the `a2a` CLI (github.com/a2aproject/a2a-cli). Use when you need to hand a task to a specialized or remote agent, fetch an agent's card, send a message to an agent at a URL, stream or poll a running task, or check, list, resume, or cancel an A2A task. Not for building or serving an A2A agent, or for calling non-A2A HTTP APIs.
Send images, audio, video, or documents into an AG2 `Agent` alongside text. Pass `ImageInput`, `AudioInput`, `VideoInput`, or `DocumentInput` as positional args to `agent.ask(...)`. Use when the user wants the agent to process non-text input — describe a photo, transcribe audio, summarise a PDF, analyse a video. Covers per-provider support matrix, the four ways to source data (URL / path / bytes / file_id), Gemini-specific YouTube + media-resolution + clipping, OpenAI image-detail, Anthropic prompt-caching on attachments, and `FilesAPI` for upload lifecycle.
Monitor an AG2 agent's stream — log events, detect repeated tool calls, track token spend, build trigger-driven observers, route observer alerts to the model, and halt on FATAL conditions. Covers `@observer(...)` (stateless), `BaseObserver` (stateful), built-ins (`TokenMonitor`, `LoopDetector`), `Watch` primitives (`EventWatch`, `CadenceWatch`, `DelayWatch`, `IntervalWatch`, `CronWatch`, `AllOf`, `AnyOf`, `Sequence`), `ObserverAlert` (`Severity.INFO/WARNING/CRITICAL/FATAL`), `AlertPolicy`, and `HaltEvent`. Use when the user wants observability, runtime safety guards, alerts, or batch/time-based reactive logic.
Get a typed Python value back from an AG2 `Agent` instead of free text. Pass `response_schema=` (a Pydantic model, dataclass, primitive, union, `ResponseSchema`, or `@response_schema` validator) and read the parsed result via `await reply.content()`. Use when the user wants validated structured output, classification, extraction, or scoring. Covers `ResponseSchema`, `@response_schema`, `PromptedSchema` (for providers without native structured output), per-turn override, validation retries, and primitive embedding.
Test AG2 agents and tools without hitting a real LLM provider. Pass `TestConfig(...)` from `ag2.testing` as the agent's config (or per-`ask`) to mock LLM responses, inject `ToolCallEvent`s to simulate tool execution, and assert success / error paths. Use when the user is writing pytest tests for an Agent or Tool.
AI perspective journaling - document daily experiences, emotions, and learnings from the agent's viewpoint. Use when asked about diary, journal entries, self-reflection, or documenting AI experiences. Creates structured daily entries capturing projects, wins, frustrations, learnings, and emotional states.
Analyze repository structure and generate or update standardized AGENTS.md files that serve as contributor guides for AI agents. Supports both single-repo and monorepo structures. Measures LOC to determine character limits and produces structured documents covering overview, optional ownership maps, patterns, conventions, and working agreements. Update mode refreshes only standard sections while preserving user-defined custom sections. Existing unmarked files require a concrete change preview and one-time user confirmation before adoption; heading matches alone never authorize rewriting. Use when the user wants an AGENTS.md generated or refreshed — setting one up for a new repository, onboarding AI agents to an existing codebase, updating an existing AGENTS.md after meaningful project changes, or when the user mentions AGENTS.md. Not for CLAUDE.md, README, or other documentation files.
Optimize agent context files (AGENTS.md, CLAUDE.md, .cursorrules, .windsurfrules, .github/copilot-instructions.md, codex.md, etc.) using Addy Osmani's agents-md methodology. Triggers: 'optimize CLAUDE.md', 'streamline CLAUDE.md', 'agents-md', 'discoverability filter', 'add gotchas', 'optimize AGENTS.md', 'optimize context file', 'CLAUDE.md 최적화', 'CLAUDE.md 줄이기', 'CLAUDE.md 다이어트', 'optimize .cursorrules', 'optimize .windsurfrules', 'optimize copilot-instructions', 'optimize codex.md'.
Use this skill ALWAYS whenever the user asks you to write code, modify files, fix bugs, plan features, or do any software development task. Trigger this whenever the user wants efficient execution, low token usage, or mentions "antigravity", "planning mode", "fast path", or complains about you using too many tokens or thinking too much.
Organize durable agent research outputs as summaries, analysis, and evidence dossiers. Use when producing multi-layer research artifacts or coordinating research handoffs. Do not use this skill for unrelated requests; route to the nearest named specialist.
Map plain-language Nexor requirements to exact configuration and build it through Nexor MCP tools. Covers arbitrary lead metadata, notifications, CRM sync, field/status events, qualification gates, workflow tools and hooks, webhooks, jobs, cloud/scheduled functions, cadence, channel and knowledge-base assignment, appointment agents (meeting types, hosts, availability, calendar connection, reminders, host routing, Calendly and external bookers), multi-agent systems, and transfers. Use when a customer asks how an agent can receive or use custom lead information, describes custom behavior, an integration, routing, meeting ownership, wants an agent that books meetings, or asks for a multi-agent build. Discover account configuration before questions, decompose variables, stages, and boundaries, split every build into modes by intent (2–3 by default, one mode only on the operator's explicit instruction, never 5+), require plan review and sign-off before mutation, prefer deterministic primitives, and read back every expected channel, connection, and knowledge assignment regardless of creation order.
Close out a finished <DOC>.final.md — fold what landed into the docs that already exist, carry what is still owed into <DOC>-V2.rev1.md, and delete the planning artifacts the tree no longer needs. Use when the user invokes /close-plan <feature>, or asks to clean up or retire a small feature's docs after execute-plan or build-feature has finished. Reviews ACCEPTANCE.md check by check before calling anything landed, then triages every phase as landed, pending, deferred or dropped. The smaller sibling of close-roadmap — works for any stem (PLAN, TECHNICAL, …); a plan usually does not earn its own folder of permanent documentation, so the default is to add a paragraph to a page that exists rather than create a new one
Close out a finished roadmap or feature execution by retiring its planning docs into canonical documentation and carrying what is still owed into the next roadmap — you are the archivist and the architect of the next round, not the coder. Use when the user invokes /close-roadmap <feature>, or asks to clean up a feature's docs, retire a roadmap, prepare for the next roadmap, or tidy docs/<feature>/ after execute-roadmap or build-feature has finished. Reviews ACCEPTANCE.md check by check, triages every item of ROADMAP.final.md or PLAN.final.md as landed, pending, deferred or dropped; moves pending work into ROADMAP-V2.rev1.md; turns landed work into decision records, architecture, data-model and product docs; writes the folder's README.md; deletes revs and execution artifacts the tree no longer needs; and fixes every inbound link. Nothing disappears unaccounted for; the user settles anything that would be dropped. For a small plan whose landed work folds into pages that already exist, use close-plan instead
Create, bulk-create, list, update, test, and delete Bolna dispositions for post-call structured extraction. Use when the user wants typed outcomes (lead_qualified, appointment_time, sentiment, churn_risk, consent_captured) pulled from every transcript and exposed in `extracted_data` on executions and webhook payloads. Covers Free Text vs Pre-defined answer types, validation (timestamp, numeric, boolean, email, regex), confidence scoring, copy-on-write semantics for shared dispositions, and the per-agent test endpoint.
Use when creating or maintaining a verification skill for a specific app or scenario; not for writing only a test plan, one-time acceptance testing, or read-only consultation. Generates scoped, executable verify-* workflows that validate real entry points and save evidence for reuse by agi-mode.
Prove and counter-prove a statement using parallel agents to eliminate confirmation bias. Use when someone says “dialectic,” “prove/disprove,” “stress test this claim,” “is this really true,” “argue both sides,” or when a statement needs objective analysis from opposing viewpoints.
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.