Agent Workflow Claude Skills
4 122 agent workflow skills from 665 authors. They configure the agents themselves: memory, prompts, context and other skills. Half of them fit into 1 844 tokens or less — that is what one costs your context window when the agent loads it. 778 ship runnable scripts rather than instructions alone. 5 of them cannot work without an MCP server, most often task. We also found 589 copies of these same skills sitting in other people's repositories — counted once here, not 589 times.
4 122 unique 665 authors 4 updated this month 506 from vendors
| SkillAlchemy — 一念落地，万象成形。输入任意想法或蒸馏目标，输出可安装的 SKILL.md。 内部编排 Lens（看清问题）和 LEAP（执行蒸馏/融合）。用户唯一入口。 Use when 用户说「蒸馏」「生成 skill」「融合」「我想做 X 但不知道从哪下手」。
| LEAP — 落地执行引擎。内含两条管线：A 分支蒸馏（从 raw data 提取 skill）、 B 分支融合（多 skill 编织为一个）。被 SkillAlchemy 编排器调用。 Use when 编排器判断需要蒸馏或融合时。
>- Universal discipline for any LM-driven loop — agent retries, plan-act-observe, multi-agent handoffs, optimiser passes, test-fix cycles. Encodes the one rule every framework reliable terminator. Termination must be provided by an explicit counter + exit predicate + stagnation signal + escalation path that live OUTSIDE the LM's control. This is a tool- level, framework-agnostic skill. It maps onto LangGraph (recursion_limit + state counter + interrupt), CrewAI (max_iter + max_rpm + human_input), Claude / OpenAI SDKs (max_iterations + tool_use_budget), DSPy (declared evaluation budget), Aider (REPL + loop, recursion limit, recursion_limit, GraphRecursionError, max iterations, max_iter, agent stuck, agent won't stop, runaway agent, ReAct loop not terminating, agent repeating itself.
>- Cross-framework enhancement overlay for choosing a multi-agent topology BEFORE writing any agent. A binary-question rubric — is single-agent + tools enough? do agents need to know about each other? does the output need one voice? — maps the answer to single-agent / supervisor / swarm / sequential / hierarchical. Activates when a coder agent is tempted to "split the work into roles" or reaches for a multi-agent framework. Encodes the *selection use multi-agent, single vs multi agent, do I need multiple agents, supervisor vs swarm, multi-agent vs single agent, agent team design.
SOP for building multi-agent systems with CrewAI — role-based collaboration, sequential/hierarchical processes, Flows, memory, delegation. Use when modeling agent teams with clear roles and task pipelines.
| Operating SOP for DSPy (Stanford NLP) — the declarative framework for "programming, not prompting" language models. "BootstrapFewShot", "GEPA", "Signatures + Modules", "teleprompter", "auto-tune prompts for a different LM", or whenever a brittle hand-crafted prompt pipeline needs to be turned into a *compiled*, measurable, swappable program. Do NOT activate for one-shot prompt tweaks, no-metric exploratory work, or pipelines where prompts must remain human-authored verbatim — use raw prompting or LangChain templates instead.
| Decision protocol for the map-reduce / dynamic fan-out pattern in LM pipelines — "given list L, run f(item) for each item in parallel, then combine". Activates when the coder agent is about to process N items with N LM calls (per-doc summarize, per-query retrieve, per-candidate rank, parallel tool fan-out). Encodes the *when*, *how many at once*, *what to do when one fails*, and *how to reduce* — not the API of any single Flow, `asyncio.gather`, `ThreadPoolExecutor`, LlamaIndex batch retrieval.
| the framework sent to the model, before changing anything else. Activate when an LM call produced an unexpected output (wrong answer, schema violation, refusal, truncation, cost spike, latency spike, infinite loop, "model got dumber after upgrade"). The skill enforces a 30-second inspect step BEFORE any prompt edit, model swap, retry, or temperature CrewAI `step_callback`, LangChain `set_debug`/`set_verbose`, Aider `/diff`+`--verbose`, raw OpenAI/Anthropic via `OPENAI_LOG=debug`/`ANTHROPIC_LOG=debug` or HTTPX event hooks. Do NOT activate for first-time prompt authoring, exploratory prompt design, or non-LM bugs.
The compile-readiness gate for prompt auto-optimization. Decide whether you have earned the right to run an optimizer (DSPy MIPROv2 / GEPA / BootstrapFewShot) before spending compute. Two preconditions only — a real metric, and enough examples for the optimizer you picked. Garbage metric in, garbage prompt out. Pick the optimizer by data scale; GEPA inverts the scale assumption (~10 examples + textual feedback).
>- Enhancement-overlay SOP for the reranker stage of a RAG pipeline — the "retrieve wide, rerank narrow" discipline. Activate when a calling agent owns a retrieval pipeline whose the context window is under pressure from too many marginal chunks. Encodes the one non- negotiable insight — a cheap bi-encoder retrieves *wide* for recall, then a more expensive cross-encoder (which reads query + document *together*) reranks *narrow* for precision; keep top-N=20-50, rerank to top-k=3-5. Covers when to add a reranker (and when not to), N-vs-k tuning, model choice (Cohere/Voyage API vs bge-reranker local vs SentenceTransformer cross-encoder), latency/cost budgeting, and the cross-framework mapping (LlamaIndex node postprocessors, LangChain ContextualCompressionRetriever, Cohere/Voyage rerank APIs, local cross-encoders). This is an ENHANCE overlay over the per- framework skills — cross-link `[[llamaindex]]` and `[[agentsop-hybrid-retrieval]]` for the
| when to /clear, when to keep context, and how to detect "context bleed" — the failure mode where stale conversation history biases the model against the current task. Surfaces a discipline that Aider (/clear), Claude Code (/clear), CrewAI (memory=False, re-instantiate), and LangGraph (new thread_id, subgraph isolation) all encode separately but none name as a skill.
>- Enhancement overlay for multi-agent / tool-using coder agents. Encodes the per-agent tool- scoping discipline that role-based frameworks (CrewAI, LangChain) document only as a every agent is a correctness and blast-radius risk. Activates when an agent system has tools AND there is more than one agent (or one agent holding many tools). Treat a tool as a capability grant; scope by least-privilege. ENHANCE overlay — read alongside [[crewai]], tools per agent, least-privilege agent, agent tool access, tool permissions, limit agent
| Tool skill for declaring reducers on LangGraph state keys so parallel writes merge instead of crashing. Activates whenever a coder agent designs a StateGraph with parallel branches, fan-out via Send, multi-agent topologies, one value per step`. Encodes the rule "every state key is single-writer or has a reducer — nothing in between."
| Enhancement-overlay decision protocol for STREAMING the output of long-running LLM / agent runs from the *backend*, not just wiring a typing animation in the UI. Activates when a coder agent must stream final tokens to a chat client, surface intermediate agent steps (which tool, which node, partial reasoning), emit custom tool-progress events, choose a transport (SSE vs WebSocket), or decide what to do when the client disconnects mid-stream. The langchain / langgraph skills mention stream modes but stop at "you can stream"; this skill encodes *what to stream, over what transport, and how to fail safely*.
Analyze, summarize, and extract insights from DeLive transcription sessions. Use when: user mentions DeLive, transcription, meeting transcripts, live captions, audio transcription, AI correction, corrected transcript, or transcript analysis; user wants to search, retrieve, summarize, correct, or process recorded transcripts; user asks about meeting notes, action items, discussion summaries, or transcript quality from DeLive. Requires DeLive app running locally with its MCP server or REST API.
为 Agent 设计的自动化浏览器 CLI 工具，也能操作 Electron 桌面应用。当需要与网站交互（包括页面导航）时使用
从 Claude Code 的 Session 解析 jsonl 对话记录，接续执行任务
编写或更新交接文档，让全新上下文的下一个 Agent 可以继续工作
此 Skill 指导 Gemini 将一个高阶的产品想法，通过结构化的分析与协作，转化为一份详尽、可执行的产品需求文档（PRD）。
自动化安装和更新来自 cafe3310 的所有 Agent Skills
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Link the current Claude Code session to a ticket (Linear, Jira, GitHub Issues, or GitHub Pull Requests) and cache its title/status in karma. Use when the user explicitly asks to link, attach, associate, or connect this session to a ticket, issue, or PR — e.g. "/link-ticket-to-session ABC-123", "link this session to LINEAR-42", "associate this work with issue #15", "attach session to PR octocat/repo#7". Do NOT auto-invoke from passing ticket-key mentions in normal conversation.
Ask Grok or Gemini via opencli (browser-session based, no API cost). Use when the user says "问 grok", "问 gemini", "ask grok", "ask gemini", "用 grok 查", "grok 怎么看", "让 gemini 分析", or wants a second opinion from Grok/Gemini without paying API tokens. Routes to opencli which drives the already-logged-in Chrome session.
Audit AND optimize a CLAUDE.md / AGENTS.md instruction file — score it against the five high-leverage patterns, flag anti-patterns, then apply approved fixes in place. Use when the user says 优化 CLAUDE.md / 优化 AGENTS.md / optimize my agent doc / 帮我改 claudemd, or after an audit when they want the fixes applied (not just reported).
Use when you want a second-opinion review via Codex CLI, cross-verification after another agent implements changes, debugging help, or alternative implementation proposals. Requires Codex CLI to be installed and authenticated.
Use when Codex feels slow, bloated, or heavy after heavy use. Provides safe session hygiene, archive strategy, and handoff discipline to keep daily Codex usage responsive and low-friction. Always pairs with comprehensive handoffs before archiving active work.
Use when you want Codex to review its own recent history (last N days or specific period) and improve its behavior. Produces minimal, high-signal updates to AGENTS.md and tiny reusable skills. The goal is long-term fluency — Codex gradually becomes better at your specific style, constraints, and workflows.
Guard long, ambiguous, or stateful AI-agent work from drift. Use when the user asks to run or continue a multi-step task, autonomous loop, bug fix, repo change, PR readiness check, compaction handoff, resume from previous context, cost-control checkpoint, or any task likely to span many tool calls, files, sessions, agents, or verification gates.
Use when coordinating complex tasks across multiple AI agents with a centralized handoff document for planning, execution tracking, and feedback fusion.
Safely diagnose and improve local network speed, latency, jitter, DNS, Wi-Fi, Ethernet, macOS network services, and bufferbloat. Use this skill whenever the user asks to optimize internet/network speed, make the network faster, diagnose slow Wi-Fi, latency, packet loss, DNS delay, unstable Codex/AI tool connectivity, or asks about the viral Codex network optimization workflow. Always protect VPN/proxy tools such as Clash Verge, Mihomo, Shadowrocket, Tailscale, V2Ray, Surge, and corporate VPNs; do not modify or disable them unless the user explicitly asks for that specific change.
回顾最近 N 天的 Claude Code 使用记录——扫描原始会话数据，按主题分组汇总"我都做了什么"，并从个人操作系统视角输出模式、风险与增删建议。当用户说 /recap、"看看我这几天做了什么"、"回顾一下我最近的会话"、"这两天我用 claude 干了啥"、"活动回顾" 时使用。
Audit and recommend improvements for a repository's agent-readable context, including AGENTS.md, CLAUDE.md, WARP.md, CONTRIBUTING.md, .agents/skills, and specs/ PRODUCT.md and TECH.md contracts. Use when asked to review, score, assess, or standardize repo instructions, agent onboarding, spec workflows, or cross-repo agent-context conventions. Use agentsmd-scaffold instead when the user wants to generate or apply root/scoped AGENTS.md files.
Audit and safely repair cross-runtime Skill ownership, loading, duplication, scopes, budgets, lifecycle, quarantine, and retirement. Usage statistics alone belong to skill-usage-stats. Use when explicitly governing the Skill ecosystem; ignore mentions/traces.
Audit, design, categorize, distribute, and measure agent skills using lessons from Anthropic's Lessons from building Claude Code: How we use skills. Use when reviewing an existing skill, deciding whether a workflow deserves a skill, planning a skill library, turning team knowledge into skills, choosing skill categories, writing trigger descriptions, designing progressive disclosure, or planning skill marketplace and usage measurement.
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Use when a skill is brittle, drifting, repeatedly failing, or needs a Reliable Skill Contract. Trigger for phrases like skill lifeguard, reliable skill, self-maintaining skill, negative examples, verification checkpoints, drift signals, replay hooks, or failure log to skill patch. Audits or patches skills so high-value workflows include explicit forbidden behaviors, checkpoints, machine-checkable done conditions, replay or smoke hooks, and drift detection.
>- 只读统计 Claude/Codex 的 Skill 使用证据并检查 agent 配置健康。仅用于明确的健康检查、配置诊断、使用排行或低使用候选请求；会识别 enabled 插件并单列批量审计读取。不要用于规范源、触发、全局/项目/profile/冷存储、投影或退役治理，改用 skill-ecosystem-doctor。忽略引用日志和相邻任务。
Route coding work so GPT-5.6 Sol remains the commander and reviewer while a separate GPT-5.6 Luna Max Codex CLI session performs concrete implementation. Use when the user asks for Sol to direct, plan, supervise, or review work done by Luna Max; when native Sol-to-Luna subagent spawning is unavailable or incompatible; or when a task needs an auditable plan, bounded worker ownership, verification, and review loop.
Use when the user explicitly asks for $threads, Codex-native subagents, 开几个子 agent, or a GitHub issue/PR queue needing parallel lanes, worktrees, review/merge gates, and closure audit. Do not use for OS/language threads, chat/email/forum threads, or Assistants product threads unless Codex workflow orchestration is explicit.
Review skill PRs with structured severity-rated feedback covering token budgets, routing conflicts, required sections, and repo conventions. WHEN: \"review skill\", \"review skill PR\", \"review skill changes\", \"check skill quality\", \"skill PR feedback\".
Query the integration-test storage account to find why a specific skill's tests are failing. Reads blob-stored test result files and surfaces error details. TRIGGERS: why is skill failing, skill test failures, debug skill tests, skill failing tests, analyze skill failures, why are tests failing for skill, skill test errors, investigate skill issues. DO NOT USE FOR: analyzing a GitHub Actions run report or comparing test runs across runs (use analyze-test-run).
Guidelines for writing Agent Skills that comply with the agentskills.io specification. WHEN: \"create a skill\", \"new skill\", \"write a skill\", \"skill template\", \"skill structure\", \"review skill\", \"skill PR\", \"skill compliance\", \"SKILL.md format\", \"skill frontmatter\", \"skill best practices\".
Author, validate, and run Vally evaluation suites for agent skills. TRIGGERS: create eval, write eval, add eval, run eval, validate eval, vally eval, eval.yaml, add stimulus, map test to eval, migrate test to eval, eval graders, eval scoring, add eval to CI.
Provision Microsoft Entra Agent Identity Blueprints, BlueprintPrincipals, and per-instance Agent Identities via Microsoft Graph, and configure OAuth 2.0 token exchange (fmi_path, OBO, cross-tenant) including the Microsoft Entra SDK for AgentID sidecar. USE FOR: Agent Identity Blueprint, BlueprintPrincipal, agent OAuth, fmi_path token exchange, agent OBO, Workload Identity Federation for agents, polyglot agent auth, Microsoft.Identity.Web.AgentIdentities. DO NOT USE FOR: standard Entra app registration (use entra-app-registration), Microsoft Foundry agent authoring (use microsoft-foundry).
Search and retrieve agent skills at runtime. This skill should be used when the agent needs to find specialized capabilities, workflows, or domain knowledge to accomplish a task. Skyll aggregates skills from skills.sh and returns full SKILL.md content ready for context injection.
> Use this skill whenever the user asks anything about Higgsfield AI — writing or refining video/image prompts, choosing a model (Kling, Sora 2, Veo, Wan, Seedance, Minimax Hailuo, DoP, Soul, Nano Banana, Seedream, Flux, GPT Image, etc.), camera controls, named motion presets, Soul ID character consistency, Cinema Studio 2.5/3.0, Vibe Motion, troubleshooting failed generations, credit optimization, Photodump, or any mention of higgsfield.ai. Also trigger on generic "write me a video prompt" or "make me an AI video prompt" requests when Higgsfield is the user's configured platform.
Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs. Use when the user wants to build a character, design a world, develop a story or premise, create a character sheet / character bible / story bible, lock a visual style or 'visual DNA', plan a multi-shot narrative with consistent characters, or says things like 'help me design a character', 'build the world', 'I need a backstory', 'make this character consistent across shots', 'develop my film/series concept', or 'I keep getting generic AI characters'. Routes the locked outputs into higgsfield-prompt + the right model. Adapted from Higgsfield's official character-design framework by @vavavinca.
Use when building, writing, refining, or structuring a Higgsfield AI prompt. Covers the MCSLA formula, prompt structure, narrative vs. timestamped formats, and how to write for both text-to-video and image-to-video workflows.