Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
End-to-end Blender 5.2 playbooks chaining the blender-* skills with QA gates - game props (mid-poly, high-to-low), game characters (sculpt/AI mesh to rigged, animated export), Roblox props/accessories/avatars, AI or downloaded asset cleanup, procedural environments, web/AR assets, product turntables, mood/weather shots (night, rain, snow, fog). Use for complete multi-stage asset requests.
Generate several genuinely different throwaway variations (designs, approaches, drafts) for the user to react to. Use when the user can only recognize what they want by seeing it — visual design, UX flows, naming, tone — or asks to brainstorm or prototype before building.
Use when a task requires interacting with a website beyond just reading it — clicking elements, filling forms, submitting data, navigating through multi-step flows, taking screenshots, or any workflow where the user needs a real browser with actions like click, type, scroll, or select. Also use for managing browser bookmarks, history, or tabs. Trigger whenever the user mentions browseros, browseros-cli, or BrowserOS. Do NOT use when simply fetching or reading page content would suffice — use curl, fetch, or WebFetch for that instead.
An end-to-end Catlass operator development orchestrator. Based on ascend-kernel (csrc/ops), it connects catlass design, catlass-operator-code-gen, and the ascendc sub-skill to complete the cycle from project initialization through documentation, accuracy, and performance. Keywords: Catlass, end-to-end, ascend-kernel, operator development, workflow orchestration.
Gmail customer support triage with mandatory full-thread read-back before every reply or follow-up, a draft-only default, per-draft Codex project threads, approval-gated final closure replies with eligible optional Trustpilot review invitations, and approval-gated operations for verified DreamCut Discord invite requests. Use when the user asks to run the customer email automation, check unread/recent support emails, review a customer follow-up, prepare or send Gmail replies, close a positively confirmed resolved case, continue a support handoff, create agent/project threads for drafted follow-up, or fulfill DreamCut Discord access requests
Compatibility router for LangWatch evaluation requests. Use only when the user asks for evaluations without making it clear whether they mean pre-deployment experiments or production online evaluations. Routes the request to the focused companion skill and does not implement either workflow itself.
Deploy AI Agent development platform (Dify) on Huawei Cloud Flexus L instance, providing deployment operations, password management, MaaS model configuration, and workflow import capabilities. Trigger keywords: deploy flexusagent/一键部署Flexus AI Agent开发平台、change password/修改开发平台管理员密码、change dify password/修改dify平台密码、add maas provider/添加MaaS模型供应商、configure maas model/配置MaaS模型、view workflow/查看AI Agent工作流、import workflow/导入AI Agent工作流
Always consult this before writing or outputting anything at the start of any communication with the user. It applies to all situations where text is written or output, such as chat responses, reports, explanations, work updates, and documentation.
Group recent conversations into overall requests, review each one's progress from memory, link each to the owner goals it serves, and write the build story, so My Work shows progress meters, shipping velocity and learned working preferences. Use when asked to "group requests", "group by overall request", "review request progress", "refresh progress", "build story", "shipping velocity", or "how do I like to work with agents".
Two-layer orchestration runtime — Main Orchestrator fetches all issues, plans execution order, maintains a rolling pool of Sub-Coordinators (up to PARALLEL_JOBS concurrently), fills freed slots immediately on completion, and updates GitHub. Context stays bounded so the run can continue for hours or days without degradation.
Design, create, run, monitor, and maintain Vapi Simulations for assistants and squads. Use for turning plain-language requirements into a coverage plan, simulation personalities, scenarios, structured-output success criteria, simulations, suites, extending an existing suite, chat or voice runs, tool mocks, target variables, lifecycle webhooks, regression coverage, launch readiness, CI quality gates, run-result analysis, and simulation API validation errors. Do not use for fixed-turn mock-conversation Evals unless the user is deciding between Evals and Simulations.
Discovers and injects project-specific coding guidelines from .trellis/spec/ before implementation begins. Reads spec indexes, pre-development checklists, and shared thinking guides for the target package. Use when starting a new coding task, before writing any code, switching to a different package, or needing to refresh project conventions and standards.
Full-process development orchestration for Ascend Triton operators. Use when users need to develop a Triton operator from scratch, follow an end-to-end development process, or are unsure which sub-skill to use. Automatically orchestrates: environment setup → requirements design → code generation → static review → accuracy validation → performance evaluation → performance optimization. Keywords: full process, development orchestration, end-to-end, workflow orchestration.
Detailed development workflow patterns, checklists, and standards. Auto-loads for complex tasks, planning, debugging, or when explicit patterns are needed. Contains session protocols, git conventions, security checklists, review-based validation, and communication standards.
Set up and improve harness engineering (AGENTS.md, docs/, lint rules, eval systems, project-level prompt engineering) for AI-agent-friendly codebases. Triggers on: new/empty project setup for AI agents, AGENTS.md or CLAUDE.md creation, harness engineering questions, making agents work better on a codebase. ALSO triggers when users are frustrated or complaining about agent quality — e.g. 'the agent keeps ignoring conventions', 'it never follows instructions', 'why does it keep doing X', 'the agent is broken' — because poor agent output almost always signals harness gaps, not model problems. Covers: context engineering, architectural constraints, multi-agent coordination, evaluation, long-running agent harness, agent loops (autonomous operation, verifiers, stop conditions), agent memory (cross-session memory, consolidation), and diagnosis of agent quality issues
Reconcile the in-memory code graph (wiki substrate) with the actual codebase. In --touched mode, re-summarizes only files touched by a completed plan (a low-cost post-plan gate). In --sweep mode, walks the entire repo and patches drift from out-of-band edits. Drift is accepted between gates; this skill is the only place where the code graph is reconciled with the repo tree.
OpenClaw learning expert that retrieves and synthesizes information from official documentation (https://docs.openclaw.ai) and GitHub repository (https://github.com/openclaw/openclaw). Use this skill whenever the user asks questions about OpenClaw, including installation, configuration, API usage, concepts, troubleshooting, best practices, or any OpenClaw-related inquiries. Triggers include OpenClaw questions about features, implementation, usage, setup, or any openclaw-related topics
Drive structured autoresearch iteration after evo:discover and the baseline commit. Use when the user invokes /evo:optimize or asks to try ideas, try variants, run experiments, use available GPUs, improve the current best/frontier, continue an evo search, or compare candidate changes in an evo workspace. The orchestrator plans and spawns optimization subagents; candidate edits/runs belong to those subagents. Width is set via subagents=N (1 for serial workloads, larger for parallel); the loop's structural value applies at any width.
Map what the owner wants from the prompts they typed themselves (prompts an agent wrote are left out), then put their top priorities and open asks, each with dated quotes, in the Goals panel at the top of My Work. Use when asked for "owner goals", "my goals", "what do I want", "what am I asking for", "map my prompts", or to refresh Goals in My Work.
Use when the developer wants their own way of working captured as a new skill - "make a skill for how we do X", a house convention or preference worth teaching every future session, a team ritual, or the same correction arriving for the second time. Interviews the developer, mines the real conventions from the database, and writes a skill in this pack's format.
Create, improve, and manage Droid skills. Use when the user wants to: - Create new skills from scratch or from session learnings - Improve existing skills based on user preferences - Analyze sessions to identify patterns worth codifying - Understand best practices for agentic skill design This is a meta-skill for self-improvement and continuous learning.
Provides comprehensive guidance for Spring AI including AI model integration, prompt templates, vector stores, and AI applications. Use when the user asks about Spring AI, needs to integrate AI models, implement RAG applications, or work with AI services in Spring.
Session context persistence for AI coding via a `.bonfire/index.md` file convention. This skill is the cross-agent fallback for hosts without a native bonfire adapter (Pi and Claude Code have first-party adapters that auto-update the file; install those when available)
You MUST use this before any creative work — creating features, building components, adding functionality, modifying behavior, or when users request help with ideation, marketing, and strategic planning. Explores user intent, requirements, and design before implementation using research-validated prompt patterns.
[AUTO-INVOKE] MUST be invoked at the START of each new coding session. Covers context management, task strategies, and Foundry-specific workflows. Trigger: beginning of any new conversation or coding session in a Solidity/Foundry project.
When the user wants to build GTM automation with code, design workflow architectures, use AI agents for GTM tasks, or implement the 'architecture over tools' principle. Also use when the user mentions 'GTM engineering,' 'GTM automation,' 'n8n,' 'Make,' 'Zapier,' 'workflow automation,' 'Clay API,' 'instruction stacks,' 'AI agents for GTM,' or 'revenue automation.' This skill covers technical GTM infrastructure from workflow design through agent orchestration.
Structured multi-angle exploration running 2-8 sequential passes over the same problem, each through a different scientific lens, to surface hidden requirements and edge cases invisible from one angle; each pass yields new ISC criteria. USE WHEN iterative depth, explore deeper, multi-angle analysis, surface hidden requirements, blind spot check, what am I missing. NOT FOR scope/zoom analysis (use ApertureOscillation).
Write explicit strategy into system prompts as a learning mechanism — making LLM behavior more predictable and teachable through prompt-as-textbook approach. Use this skill when the user wants to improve an agent or LLM's performance on a specific task through system prompts, needs to encode a strategy or workflow into a prompt, wants consistent LLM behavior, or says "improve system prompt", "encode this strategy", "prompt engineering", "make agent follow this workflow", "prompt as textbook". Based on Karpathy system prompt learning post.
Use when parsing agent log files to identify error patterns, rate limit hits, timeout clusters, tool failures, and component-level error counts. Produces a structured anomaly report. Cron-compatible — silent if no issues, alert digest if anomalies found.
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.