智能体
Agent 规划、多步流程与编排
mastering-langgraph
SpillwaveSolutions
Build stateful AI agents and agentic workflows with LangGraph in Python. Covers tool-using agents with LLM-tool loops, branching workflows, conversation memory, human-in-the-loop oversight, and production monitoring. Use when - (1) building agents that use tools and loop until task complete, (2) creating multi-step workflows with conditional branches, (3) adding persistence/memory across turns with checkpointers, (4) implementing human approval with interrupt(), (5) debugging via time-travel or LangSmith. Covers StateGraph, nodes, edges, add_conditional_edges, MessagesState, thread_id, Command objects, and ToolMessage handling. Examples include chatbots, calculator agents, and structured workflows.
tensorlake
tensorlakeai
Tensorlake SDK for agent sandboxes and sandbox-native orchestration. Use when the user mentions tensorlake, or asks about Tensorlake APIs/docs/capabilities. Also use when the user is building AI agents or agentic applications that need stateful sandboxed execution environments for agents and isolated tool calls, with suspend/resume and snapshots for persistence between tasks, or durable workflow orchestration for agents (parallel map/reduce DAGs). Works with any LLM provider (OpenAI, Anthropic), agent framework (LangChain), database, or API as the infrastructure layer.
bug-hunt-swarm
Dimillian
"Parallel read-only multi-agent root-cause investigation for bugs, regressions, crashes, flaky behavior, or unexplained failures. Use when the user asks to investigate a bug, find the root cause, trace a regression, understand why something broke, or wants a ranked diagnosis with the fastest proof path without making code edits."
feature-implementation
NickCrew
Use when implementing a feature or multi-file code change - provides structured implementation flow with persona selection, validation, and testing guidance.
ms
tkersey
Create, update, or refactor Codex skills in this repo, including SKILL.md, frontmatter trigger text, agents/openai.yaml, and scripts/references/assets. Use when asked to create or improve a skill, tighten trigger descriptions from usage evidence (for example via $seq), regenerate/verify agents/openai.yaml, or adjust skill metadata/workflows.
Parallelization Patterns
takeokunn
Patterns for parallel execution strategies, timeout configuration, retry policies, and consensus mechanisms.
Define Core
takeokunn
Shared workflow phases and patterns for requirements definition commands. Use this skill when implementing /define or /define-full commands to ensure consistent workflow structure, agent delegation, and requirements documentation patterns.
agent-evaluation
mlflow
Use this when you need to EVALUATE OR IMPROVE or OPTIMIZE an existing LLM agent's output quality - including improving tool selection accuracy, answer quality, reducing costs, or fixing issues where the agent gives wrong/incomplete responses. Evaluates agents systematically using MLflow evaluation with datasets, scorers, and tracing. Covers end-to-end evaluation workflow or individual components (tracing setup, dataset creation, scorer definition, evaluation execution).
searching-mlflow-docs
mlflow
Searches and retrieves MLflow documentation from the official docs site. Use when the user asks about MLflow features, APIs, integrations (LangGraph, LangChain, OpenAI, etc.), tracing, tracking, or requests to look up MLflow documentation. Triggers on "how do I use MLflow with X", "find MLflow docs for Y", "MLflow API for Z".
agent-teams-playbook
KimYx0207
Cross-runtime Agent Teams orchestration playbook for Claude Code, Codex, OpenClaw, and Cursor. This skill should be used when the user asks to "create agent teams", "use agent swarm", "setup multi-agent collaboration", "orchestrate agents", "coordinate parallel agents", "organize team collaboration", "build agent teams", "implement swarm orchestration", "setup multi-agent system", "coordinate agent collaboration", or needs guidance on adaptive team formation, quality gates, skill discovery, task distribution, team coordination strategies, or Agent Teams best practices. 或者当用户说"多agent"、"agent协作"、"agent编排"、"并行agent"、"分工协作"、"拉团队"、"拉个团队"、"多代理协作"、"swarm编排"、"agent团队"时也应使用此技能。Note: "swarm/蜂群" is a generic industry term; Claude Code's official concept is "Agent Teams"; Codex, OpenClaw, and Cursor should map this playbook to their native or host-provided agent capabilities without deleting any workflow stage.
setup-ralph
glittercowboy
Set up and configure Geoffrey Huntley's original Ralph Wiggum autonomous coding loop in any directory with proper structure, prompts, and backpressure.
create-subagents
glittercowboy
Expert guidance for creating, building, and using Claude Code subagents and the Task tool. Use when working with subagents, setting up agent configurations, understanding how agents work, or using the Task tool to launch specialized agents.
create-agent-skills
glittercowboy
Expert guidance for creating, writing, building, and refining Claude Code Skills. Use when working with SKILL.md files, authoring new skills, improving existing skills, or understanding skill structure and best practices.
cc10x-router
romiluz13
THE ONLY ENTRY POINT FOR CC10X. This skill MUST be activated for ANY development task - never skip. Use this skill when: building, implementing, debugging, fixing, reviewing, planning, refactoring, testing, or ANY coding request. If user asks to write code, fix bugs, review code, or plan features - USE THIS SKILL. Triggers: build, implement, create, make, write, add, develop, code, feature, component, app, application, review, audit, check, analyze, debug, fix, error, bug, broken, troubleshoot, plan, design, architect, roadmap, strategy, memory, session, context, save, load, test, tdd, frontend, ui, backend, api, pattern, refactor, optimize, improve, enhance, update, modify, change, help, assist, work, start, begin, continue, research, cc10x, c10x. CRITICAL: Execute workflow immediately. Never just describe capabilities.
takopi-release
banteg
Prepare and ship a Takopi release. Use when asked to cut a release, bump release versions, update changelog/spec/readme, tag v<major.minor.patch>, or trigger the GitHub release workflow.
meta-audit
laurigates
Audit Claude subagent configurations for completeness, security, and best practices
business-analyst
aj-geddes
Product discovery and requirements analysis specialist. Conducts stakeholder interviews, market research, problem discovery, and creates product briefs. Use for product brief, brainstorm, research, discovery, requirements gathering, problem analysis, user needs, competitive analysis, and setting foundation before product planning. Hands off to product manager when analysis complete.
product-manager
aj-geddes
Product requirements and planning specialist. Creates PRDs and tech specs with functional/non-functional requirements, prioritizes features using MoSCoW/RICE frameworks, breaks down epics into user stories, and ensures requirements are testable and traceable. Use for PRD creation, requirements definition, feature prioritization, tech specs, epics, user stories, and acceptance criteria.
workflow-automation
JoelLewis
"Workflow automation for securities operations: BPM concepts, task routing, approval chains, escalation rules, SLA monitoring, and process orchestration patterns."
bmad-orchestrator
aj-geddes
Orchestrates BMAD workflows for structured AI-driven development. Use when initializing BMAD in projects, checking workflow status, or routing between 4 phases (Analysis, Planning, Solutioning, Implementation). Manages project configs, tracks progress through project levels 0-4, and coordinates with specialized workflows. Trigger on /workflow-init, /workflow-status, or when users need BMAD setup.
creative-intelligence
aj-geddes
Facilitates structured brainstorming sessions, conducts comprehensive research, and generates creative solutions using proven frameworks. Trigger keywords - brainstorm, ideate, research, SCAMPER, SWOT, mind map, creative, explore ideas, market research, competitive analysis, innovation, problem solving, feature generation
philosophy-of-work
atrislabs
"Install Atris — the open-source operating system for AI work. Gives agents MAP-first navigation, durable tasks, proof on disk, and a plan→do→review loop. Use when starting in a repo, onboarding, or when chat context keeps dying."
autopilot
atrislabs
Run one bounded Atris autonomy step. Use when the user says autopilot, run one tick, do the next thing, keep going, or make progress without losing proof.
endgame
atrislabs
"Backward partner to /flow. Picks the next horizon from inbox + wiki + lessons, writes the reverse path as tagged tasks to TODO.md so /autopilot can pursue it tick by tick. Triggers on: endgame, what's the last move, where are we heading, reverse engineer, work backward."