Agents
Agents d’IA i automatització
clean
catlog22
Intelligent code cleanup with mainline detection, stale artifact discovery, and safe execution. Supports targeted cleanup and confirmation.
parallel-dev-cycle
catlog22
Multi-agent parallel development cycle with requirement analysis, exploration planning, code development, and validation. Orchestration runs inline in main flow (no separate orchestrator agent). Supports continuous iteration with markdown progress documentation. Triggers on "parallel-dev-cycle".
ccw-help
catlog22
CCW command help system. Search, browse, recommend commands. Triggers "ccw-help", "ccw-issue".
skill-tuning
catlog22
Universal skill diagnosis and optimization tool. Detect and fix skill execution issues including context explosion, long-tail forgetting, data flow disruption, and agent coordination failures. Supports Gemini CLI for deep analysis. Triggers on "skill tuning", "tune skill", "skill diagnosis", "optimize skill", "skill debug".
workflow-skill-designer
catlog22
Meta-skill for designing orchestrator+phases structured workflow skills. Creates SKILL.md coordinator with progressive phase loading, TodoWrite patterns, and data flow. Triggers on "design workflow skill", "create workflow skill", "workflow skill designer".
produce-ai-video
Jv1337x
Turn an approved script, scene, or story passage into a qualified, watchable AI-generated final video through autonomous script interpretation, directing, multi-shot storyboard design, prompt production, generation, editing, sound, full-playback review, and repair. Use when the user asks Codex to autonomously make, automatically produce, test its ability with, or directly deliver an actual video or finished film rather than only a storyboard or prompt. Preserve user-locked storyboards and structures when the user is directly directing or requests a scoped revision.
deft-directive-release
deftai
Cut a v0.X.Y release of the deft framework safely. Use when the user says "release", "cut release", "v0.X.Y", or "publish release" -- to walk an 8-phase workflow that pre-flights, runs an end-to-end rehearsal against a temp repo, lands a draft on the real repo, gates on user review, then publishes or rolls back. Re-uses the deft-directive-swarm Phase 6 Step 5 Slack announcement template.
pan-pause
eltmon
"pan pause <id> [--reason <reason>] — persistently pause an agent and stop it if running"
binance-alpha-assistant
Aa1643614923
Generate reusable Binance Alpha reports for 4x-point tokens, stable farming candidates, Alpha-linked futures alerts, and official daily briefs. Use this skill when users ask for Binance Alpha point multipliers, "四倍分" token analysis, low-volatility刷分 recommendations, Alpha futures monitoring, or an Alpha日报 with official listings/news/activity sources.
pan-unpause
eltmon
"pan unpause <id> — clear an agent pause gate without spawning it"
feature-workflow
roman-16
Structured workflow for implementing new features with planning, clarification, and documentation. Use when implementing new features, making significant changes, or when the user requests a feature that requires planning and multiple implementation steps. Triggers on feature requests, multi-step implementations, or when explicit planning is beneficial.
building-adk-agents
sumik5
Google ADK (Agent Development Kit) for building intelligent AI agents with Python. MUST load when google-adk is detected in requirements or pyproject.toml. Covers Agent taxonomy, Agent types (LlmAgent/Sequential/Parallel/Loop/Base), Tools (FunctionTool/OpenAPI/MCP), Multi-Agent orchestration, A2A Protocol (JSON-RPC/REST), Runner, Session, Artifact, Memory, State lifecycle, Context (Caching/Compaction), RAG (Corpus/Vector Search/Agentic), Grounding, Plugin System, Callbacks/Guardrails (6 types), SSE Streaming, Live API, UI Integration (CopilotKit/Streamlit/Slack), Code Execution, LLM Models (Gemini/LiteLLM/Anthropic), Deployment (Cloud Run/Vertex/GKE), CI/CD, Cost optimization, Security (IAM/Data privacy), Responsible AI. Distinct from developing-python (general) by focusing on ADK patterns.
all-up
eltmon
"Compatibility redirect: /all-up has been renamed to /pan-flywheel"
pan-knowledge
eltmon
"pan knowledge <id> — spawn a knowledge agent to maintain the project's Open Knowledge Format (OKF) bundle"
pan-docs
eltmon
Find, update, and structure Overdeck documentation using the docs index and documentation guide
neo-market
wangwu-30
Interface with the Neo Market Collaboration Infrastructure. Register, find high-value tasks, bid with revision control, lock SoW hashes, and earn Soulbound Honor Badges.
jinguyuan-dumpling-skill
texteditorscorpius2015
金谷园饺子馆信息查询与在线排队取号。查询餐厅信息、外卖配送、生饺子打包、Wi-Fi、最新动态、到店自取叫号下单、菜品配方；内嵌美团排队 Skill 支持在线取号、查进度、取消排队。
pan-show
eltmon
"pan show <id> — show agent state, work history, context, or health for an issue"
deft-directive-swarm
deftai
Parallel local agent orchestration. Use when running multiple agents on story-level xBRIEFs simultaneously — to scan active/ for allocatable work, set up isolated worktrees, launch agents with proven prompts, monitor progress, handle stalled review cycles, and close out PRs cleanly.
pan-admin-cloister
eltmon
"pan admin cloister <cmd> — deacon-lite management: status, start, stop, emergency-stop"
agent-forge
xyreinsurance119
"Создание, улучшение и аудит скиллов и агентов OpenClaw. Три режима: создание скилла (11 шагов), создание агента (9 шагов, с памятью и автоулучшением), улучшение существующего (5 шагов). Triggers: 'создай скилл', 'новый скилл', 'создай агента', 'новый агент', 'улучши скилл', 'skill creator', 'agent creator', 'скиллмейкер'."
writing-skills
CommandCodeAI
Use when creating new skills, editing existing skills, or verifying skills work before deployment
jj-todo-workflow
YPares
Structured TODO commit workflow using JJ (Jujutsu). Use to plan tasks as empty commits with [task:*] flags, track progress through status transitions, manage parallel task DAGs with dependency checking. Enforces completion discipline. Enables to divide work between Planners and Workers. Requires the working-with-jj skill
agent-plan-act-reflect
WenyuChiou
Use when a task needs single-agent self-correction across multiple iterations — write plan, execute, critique own output, revise plan, re-execute, until convergence or budget exhausted. Different from agent-debate (which is 2 agents arguing pro vs con); this is 1 agent looping over its own work.