agentsclimarketplace

Context & AI engineering

25,745 rows, by stacks then stars

1,865 of these skill files have been read, by 1,328 distinct authors, and what they tell an agent counted. What Context & AI engineering authors agree on, and what they forbid

  • Glb animation handoff

    ThanhNguyxnOrg/blendops/skills/glb-animation-handoff Skill

    4★ repo

    Document GLB animation contract (skeletal vs morph, FPS, looping, root motion, blend transitions) with truthful caveats for web playback.

  • Glb web handoff

    ThanhNguyxnOrg/blendops/skills/glb-web-handoff Skill

    4★ repo

    Prepare truthful GLB/web handoff output with explicit artifact status, assumptions, and caveats.

  • Pre handoff verification

    ThanhNguyxnOrg/blendops/skills/pre-handoff-verification Skill

    4★ repo

    Verification gate before claiming any deliverable is ready. Forces evidence checks (path, file existence, validation notes) per evidence-before-done law. Use before any 'Verified' claim.

  • Po ui

    tbc-servicos/dataagile-agent-kit/po-ui/skills/po-ui Skill

    4★ repo

    Consulta componentes, inputs/outputs, exemplos e guias da biblioteca PO-UI (Angular TOTVS) via MCP server oficial. Use antes de escrever ou modificar qualquer UI que utilize PO-UI.

  • Lo handoff manager

    stephenrogan/augmented-csm/skills/pillar-2-lifecycle-orchestration/lo-handoff-manager Skill

    4★ repo

    Generates handoff documentation and manages transition sequences when accounts move between teams, CSMs, or lifecycle stages. Covers sales-to-CS handoff, CSM-to-CSM reassignment, CS-to-renewal team handoff, and any structured account transition. Use when asked to create a handoff brief, manage an account transition, prepare for CSM reassignment, build a sales-to-CS handoff, transfer account knowledge, or when any account is changing ownership. Also triggers for questions about knowledge transfer, account reassignment, team transitions, or onboarding a new CSM onto an existing account.

  • Lo onboarding orchestrator

    stephenrogan/augmented-csm/skills/pillar-2-lifecycle-orchestration/lo-onboarding-orchestrator Skill

    4★ repo

    Manages the customer onboarding workflow from sales handoff through first-value milestone. Sequences tasks, triggers communications, tracks completion, and flags delays. Use when asked to set up onboarding for a new customer, track onboarding progress, build onboarding workflows, automate onboarding sequences, manage customer go-live, monitor time-to-value, or when any new account needs to be moved through a structured activation process. Also triggers for questions about onboarding playbooks, kickoff coordination, or implementation tracking.

  • Lo qbr orchestrator

    stephenrogan/augmented-csm/skills/pillar-2-lifecycle-orchestration/lo-qbr-orchestrator Skill

    4★ repo

    Manages the end-to-end QBR process including scheduling, attendee coordination, content preparation, and follow-up tracking. Use when asked to schedule a QBR, prepare a business review, build QBR content, automate QBR preparation, coordinate a quarterly review, generate a QBR deck, or when any account needs a structured business review. Also triggers for questions about business review cadence, executive review preparation, or customer success review meetings.

  • Pa benchmark engine

    stephenrogan/augmented-csm/skills/pillar-4-product-adoption/pa-benchmark-engine Skill

    4★ repo

    Benchmarks account adoption metrics against segment peers to show customers where they stand relative to similar companies. Produces comparison data for QBRs, adoption conversations, and value narratives. Use when asked to benchmark a customer against peers, create adoption comparisons, show how an account compares to similar companies, build segment comparison data, or when any customer-facing conversation needs peer benchmarking context. Also triggers for questions about peer comparison, industry benchmarks, best-in-class adoption metrics, or competitive positioning within a segment.

  • Pa enablement orchestrator

    stephenrogan/augmented-csm/skills/pillar-4-product-adoption/pa-enablement-orchestrator Skill

    4★ repo

    Coordinates customer enablement activities including training session scheduling, learning path management, certification tracking, and enablement content recommendation based on adoption gaps. Bridges the gap between product access and product competence. Use when asked to schedule training, manage customer education, track certification progress, recommend learning resources, build an enablement plan, coordinate onboarding training, or when any adoption gap could be addressed through customer education. Also triggers for questions about customer training, learning management, enablement programmes, knowledge transfer, or adoption through education.

  • Pa knowledge base curator

    stephenrogan/augmented-csm/skills/pillar-4-product-adoption/pa-knowledge-base-curator Skill

    4★ repo

    Keeps customer-facing help content aligned with product changes, identifies content gaps from support ticket patterns, surfaces relevant resources to customers based on adoption profiles, and measures documentation effectiveness through ticket deflection. Use when asked to audit help content, identify knowledge base gaps, recommend documentation improvements, match customers to relevant help articles, track documentation quality, or when support ticket patterns suggest missing or outdated documentation. Also triggers for questions about help centre management, documentation quality, self-serve content strategy, support deflection through better documentation, or content gap analysis.

  • Building mcp servers

    celigo/ai/skills/building-mcp-servers Skill

    4★ repo

    Build Celigo MCP server resources -- endpoints that expose Tools and builder-mode APIs to external AI agents and MCP clients. Use when creating MCP servers, linking tools or APIs, or configuring annotations and overrides.

  • Configuring ai agents

    celigo/ai/skills/configuring-ai-agents Skill

    4★ repo

    Configure Celigo AI agent and guardrail imports -- LLM-powered steps that classify, extract, validate, or generate data within flows. Use when creating agent imports (OpenAI, Gemini), guardrails (PII, moderation), or configuring prompts, structured output, or BYOK connections.

  • Managing on premise agents

    celigo/ai/skills/managing-on-premise-agents Skill

    4★ repo

    Manage Celigo on-premise agents -- lightweight software you install on a host inside your own network that opens a secure OUTBOUND tunnel to integrator.io, letting the platform reach private systems (on-prem databases, file shares, internal apps) without opening inbound firewall ports. An on-premise agent is a connectivity resource on the `/agents` page -- not a stack (compute) and not an AI agent (LLM processing). Use when connecting to a system behind a firewall, installing or pairing an agent, rotating its access token, or diagnosing agent-offline flow failures.

  • Releases mcp

    buildinternet/releases-cli/skills/releases-mcp Skill

    4★ repo

    Use when the user asks about recent releases, changelogs, what's new in a library, breaking changes, version updates, or wants to compare products. Activates for questions like "what changed in Next.js 15?", "latest Tailwind releases", "compare Bun vs Deno releases".

  • Zai web search

    Axect/skills/zai-web-search Skill

    4★ repo

    Search the live web via z.ai web_search_prime (GLM Coding Plan MCP server). Use when the user wants current information, recent trends, news, blog posts, docs, or any web resource that academic-database search (reference-search) does not cover. Triggers on web search, 웹 검색, search the web, 최근 동향, 최신 트렌드, latest trends, find online, 인터넷에서 찾아, what's new, current state of X, search z.ai, GLM web search.

  • Skill creator

    zebbern/icron/icron/skills/skill-creator Skill

    4★ repo

    Create or update AgentSkills. Use when designing, structuring, or packaging skills with scripts, references, and assets.

    The same file is in 1 other repository

  • Make plan

    fabienjuif/agentic-skills/plugins/fabien-skills/skills/make-plan Skill

    3★ repo

    Produce a plan for a non-trivial task as a self-contained docs/plans/<name>/ directory (CLAUDE.md, README.md, ROADMAP.md, PROMPT.md, changelogs/) so a fresh agent session AND a human can pick it up later. Use whenever asked to "create a plan", "plan this out", or before implementing anything multi-step or architectural.

  • Claude mail bridge wakeup

    CodingJay-1/claude-mail-bridge/src/claude_mail_bridge/assets/skills/claude-mail-bridge-wakeup Skill

    3★ repo

    Use when AI receives messages starting with `[Remote System : type]`. Provides guidelines for handling system wake-up messages and email communication.

  • Llm evaluation

    carlopezzuto/agents/.claude/skills/llm-evaluation Skill

    3★ repo

    Implement comprehensive evaluation strategies for LLM applications using automated metrics, human feedback, and benchmarking. Use when testing LLM performance, measuring AI application quality, or establishing evaluation frameworks.

  • Prompt engineering patterns

    carlopezzuto/agents/.claude/skills/prompt-engineering-patterns Skill

    3★ repo

    Master advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability in production. Use when optimizing prompts, improving LLM outputs, or designing production prompt templates.

  • Rag implementation

    carlopezzuto/agents/.claude/skills/rag-implementation Skill

    3★ repo

    Build Retrieval-Augmented Generation (RAG) systems for LLM applications with vector databases and semantic search. Use when implementing knowledge-grounded AI, building document Q&A systems, or integrating LLMs with external knowledge bases.

  • Vigil

    JakubKontra/skills/vigil Skill

    3★ repo

    macOS-native self-supervising watchdog that monitors a project autonomously in the background. Launch it once and it keeps watching after you close the terminal, sleep, or restart — a launchd agent periodically runs read-only health probes (tests, build, dependency audit, a localhost health URL, git drift, disk space) and fires native macOS notifications plus an optional spoken summary when something changes or breaks. Read-only by default with code-enforced safety rails; no LLM runs in the watch loop. Use when the user wants to "watch", "monitor", or "keep an eye on" their project, asks to "notify me when" tests break or the build fails, wants a "self-watching" or "macOS background" monitor or an overnight/unattended health guardian, or wants to install, check on, triage, or stop a background watcher. Has a second mode — "task completion": give a plain-language task and Vigil periodically uses a read-only AI judge to check whether it's done, then notifies you and uninstalls itself (one-shot). Use task mode when the user says "tell me when this task is done", "check if X is finished", "watch until the feature is complete", or "let me know when the work is done". Configured via vigil.config.json in the project root.

  • Skill creator

    QBall-Inc/clear/skills/skill-creator Skill

    3★ repo

    Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

  • Nudge

    voxpelli/vp-claude/skills/nudge Skill

    3★ repo

    This skill (explicit /nudge only) manages Claude Code feature-adoption nudges sourced from the Basic Memory note main/reference/claude-code-noteworthy-features. Bare /nudge = Mode A (sync): re-read the note, filter out features already marked adopted or declined, and regenerate the tip cache ~/.claude/references/claude-code-nudge-tips.txt that the SessionStart hook reads. /nudge check = Mode B (adoption check): scan Claude Code session transcripts across all projects for real evidence of feature use, preview proposed adoption-status transitions, write approved changes to the note's frontmatter, then regenerate the same cache. No delegate agent exists (unlike knowledge-garden/knowledge-maintain, whose graph-wide phrasing routes through their agent) — natural-language phrasing like 'sync nudge tips' or 'nudge me on unused features' does NOT auto-invoke this skill; recognize the intent and suggest the user run `/nudge` or `/nudge check` explicitly.

  • Onboard session

    TimboGP/timbogp-marketplace/plugin/agentic-study-environment/skills/onboard-session Skill

    3★ repo

    Begin a bracketed onboarding session over an existing artifact you didn't author — a codebase, a document corpus, a body of papers — inside a agentic-study-environment sub-project. The agent surveys the artifact, walks it part by part with comprehension checks, then has the user reproduce a recent change from its history. Use whenever the user signals they want to get up to speed on something someone else made — phrases like "onboard me on this codebase", "walk me through this repo", "help me get up to speed on X", "I just joined this project", "drop me into this codebase", or "get familiar with this corpus" should trigger this skill. `onboarding` is the fourth core session type; unlike `theory`/`practice`/`role-play` it has its own dedicated entry point rather than being proposed inside `start-session`. A session ends with the `stop-session` skill.

  • Kanso context

    blakecyze/kanso/skills/kanso-context Skill

    3★ repo

    Use when the user asks to update, prune, audit, or create AGENTS.md, CLAUDE.md, or other agent context files. Also use when the user says their AI agents keep making the same mistakes and they want to fix it at the context level.

  • Kanso prompting

    blakecyze/kanso/skills/kanso-prompting Skill

    3★ repo

    Use when rewriting, sharpening, or producing a prompt for Claude or another current frontier model. Sets standing rules for getting better LLM output through better prompts. Loaded by /kanso-task; not directly invoked by the user.

  • Templateclaw setup

    jeromwolf/templateclaw/skills/templateclaw-setup Skill

    3★ repo

    Generate project setup files — README.md, CLAUDE.md, AGENTS.md, .env template, Docker Compose. Best practices for modern development teams and AI-assisted coding.

  • Rag

    bethanychamberlain/claude-skills-librechat/librechat-data/skills/rag Skill

    3★ repo

    Use when setting up LibreChat RAG (Retrieval-Augmented Generation), configuring embeddings providers, setting up file search in agents, configuring PGVector/PostgreSQL for vector storage, or troubleshooting document indexing and retrieval. Also use when users ask about 'chat with documents' or 'file search' features.

  • Mcp

    bethanychamberlain/claude-skills-librechat/librechat-mcp/skills/mcp Skill

    3★ repo

    Use when configuring MCP (Model Context Protocol) servers in LibreChat, connecting external tools via MCP, setting up mcpServers in librechat.yaml, or debugging MCP tool connections. Also use when asked about extending LibreChat agents with external capabilities through MCP.

  • Maintaining agents md

    qte77/claude-code-plugins/plugins/docs-governance/skills/maintaining-agents-md Skill

    3★ repo

    Maintains AGENTS.md, AGENT_LEARNINGS.md, AGENT_REQUESTS.md, and CONTRIBUTING.md governance files in sync with codebase changes. Use when updating governance files, during sprint reviews, or after structural changes.

    The same file is in 1 other repository

  • Implementing document indexing

    qte77/claude-code-plugins/plugins/rag-core/skills/implementing-document-indexing Skill

    3★ repo

    Implements document indexing with heading-boundary chunking, embedding, FAISS vector store, and PageIndex-style hybrid retrieval. Use when building RAG pipelines, document search, or memory layers.

  • Workflow skill evals

    bitwise-media-group/skills/plugins/workflow/skills/workflow-skill-evals Skill

    3★ repo

    Write the evolve evaluation suite for an agent skill: its triggers (triggers.json — Tier 1 activation tests) and behavioral evals (evals.json — Tier 2 task tests), under evals/<skill>/. Use when asked to generate, write, author, scaffold, or balance an eval suite, evals, or triggers for a skill; to create or edit a triggers.json or evals.json; to add behavioral evals to a SKILL.md; to add or rebalance positive and near-miss negative trigger cases; or to measure or evaluate whether a skill activates and fires on the right prompts and does its job. Follows the evolve evaluations guide and the JSON Schemas it links, fetched at author time so it tracks new assertion types and fields. Prefers deterministic assertions (file_exists, regex, command, tool_call) over the LLM judge. Not for running or sweeping existing suites, writing application unit tests, or comparing model quality.

  • Benchmark

    CODE-SAURABH/OpenSkills/benchmark Skill

    3★ repo

    Performance benchmarking skill — baseline and compare page load times, Core Web Vitals, bundle sizes, and API response times. Use when the user wants to measure performance before and after a change, catch performance regressions in a PR, establish a performance baseline, or understand what is making their app slow.

  • Mcp builder

    CODE-SAURABH/OpenSkills/mcp-builder Skill

    3★ repo

    Build production-grade MCP (Model Context Protocol) servers that connect AI agents to external APIs, databases, and services. Use when the user wants to build an MCP server, integrate an external API with Claude or any AI agent, create custom tools for an AI agent, expose a service via MCP, or asks how to connect their app to an AI model using the MCP standard.

  • Prompt engineer

    CODE-SAURABH/OpenSkills/prompt-engineer Skill

    3★ repo

    Design, iterate, test, and version prompts for LLMs — system prompts, few-shot examples, chain-of-thought, structured output, and agentic tool-use prompts. Use when the user asks to write a prompt, improve an existing prompt, design a system prompt, add few-shot examples, control LLM output format, reduce hallucinations, or get better results from any AI model.

  • Research agent

    CODE-SAURABH/OpenSkills/research-agent Skill

    3★ repo

    49 production-grade AI agent skills (SKILL.md) for Claude Code, Codex & Antigravity — system design, DevOps, security, QA, and more. MIT licensed, open source.

  • Skill creator

    CODE-SAURABH/OpenSkills/skill-creator Skill

    3★ repo

    Create new skills, improve existing skills, and test skill quality. Use when the user wants to build a skill from scratch, turn a workflow into a reusable skill, improve or refactor an existing SKILL.md, test how well a skill performs, optimize a skill's description so it triggers correctly, or asks "how do I make a skill for X". This is the meta-skill — it builds all other skills.

  • Claude fix unicode decode

    hachiwar/codex-skills/claude-fix-unicode-decode Skill

    no license3★ repo

    Diagnose and fix UnicodeDecodeError, UnicodeEncodeError, mojibake, and corrupted Chinese text in Claude Code tasks. Use this skill whenever a file, command, test, CSV, JSON, HTTP response, Git diff, or Windows terminal shows broken Chinese, replacement characters, strings such as "中文" or "涓枃", an unknown codec, or any decode/encode exception. Use it even when the user only says Chinese text is garbled or a file cannot be read.

  • Prompt structure

    soden46/syarif-laravel-ai-skills/skills/prompt-structure Skill

    3★ repo

    Laravel guidance to structure prompts for clarity-separate concerns, prioritize requests, specify acceptance criteria, and break work into testable increments

  • Vector search

    soden46/syarif-laravel-ai-skills/skills/vector-search Skill

    3★ repo

    Add semantic search with native vector queries (Laravel 13+); pgvector similarity clauses, embedding workflows, and hybrid search patterns

  • Team

    xWael9/ClaudeBoss/skills/team Skill

    3★ repo

    Delegate a task to a free opencode agent, wait for its output, then review and implement it together with the user. Use when the user says "team mode", "/claudeboss:team", or wants Claude Code and opencode working as a pair on a task.

  • Aris research lit

    JZKK720/cubecloud-skilldbundle-setup/upstream/aris-research-lit Skill

    no license3★ repo

    Search and analyze research papers, find related work, summarize key ideas. Use when user says "find papers", "related work", "literature review", "what does this paper say", or needs to understand academic papers. Ported from wanshuiyin/Auto-claude-code-research-in-sleep (ARIS) — methodology only, Claude-Code-specific MCP/CLI deps stripped for VS Code Copilot.

  • Content loop orchestrator

    genfeedai/skills/bundles/all/skills/content-loop-orchestrator Skill

    3★ repo

    Operate the executable Genfeed content loop by routing stages across skills and automating sense and measure with the loop driver. Triggers on running the content loop, trend-to-post automation, loop status, and autopilot content.

  • Image prompt engineer

    genfeedai/skills/bundles/all/skills/image-prompt-engineer Skill

    3★ repo

    Craft optimized prompts for AI image and short video generation across major models, visual formats, aspect ratios, and content goals. Triggers on image prompts, video prompts, prompt optimization, Flux, DALL-E, Midjourney, and Genfeed visual prompts.

  • Content loop orchestrator

    genfeedai/skills/bundles/content-loop/skills/content-loop-orchestrator Skill

    3★ repo

    Operate the executable Genfeed content loop by routing stages across skills and automating sense and measure with the loop driver. Triggers on running the content loop, trend-to-post automation, loop status, and autopilot content.

  • Image prompt engineer

    genfeedai/skills/bundles/content-loop/skills/image-prompt-engineer Skill

    3★ repo

    Craft optimized prompts for AI image and short video generation across major models, visual formats, aspect ratios, and content goals. Triggers on image prompts, video prompts, prompt optimization, Flux, DALL-E, Midjourney, and Genfeed visual prompts.

  • Image prompt engineer

    genfeedai/skills/bundles/image-visual/skills/image-prompt-engineer Skill

    3★ repo

    Craft optimized prompts for AI image and short video generation across major models, visual formats, aspect ratios, and content goals. Triggers on image prompts, video prompts, prompt optimization, Flux, DALL-E, Midjourney, and Genfeed visual prompts.

  • Content loop orchestrator

    genfeedai/skills/bundles/platform-development/skills/content-loop-orchestrator Skill

    3★ repo

    Operate the executable Genfeed content loop by routing stages across skills and automating sense and measure with the loop driver. Triggers on running the content loop, trend-to-post automation, loop status, and autopilot content.

  • Content loop orchestrator

    genfeedai/skills/content-loop-orchestrator Skill

    3★ repo

    Operate the executable Genfeed content loop by routing stages across skills and automating sense and measure with the loop driver. Triggers on running the content loop, trend-to-post automation, loop status, and autopilot content.

  • Image prompt engineer

    genfeedai/skills/image-prompt-engineer Skill

    3★ repo

    Craft optimized prompts for AI image and short video generation across major models, visual formats, aspect ratios, and content goals. Triggers on image prompts, video prompts, prompt optimization, Flux, DALL-E, Midjourney, and Genfeed visual prompts.

  • Ai prompt ux

    varunk130/ai-ux-skill-library/skills/ai-prompt-ux Skill

    3★ repo

    Design the input experience for AI products - how users craft, structure, and refine their instructions to AI systems. Use when: prompt interface, AI input design, prompt templates, prompt suggestions, context window UX, instruction design, AI input affordances, prompt engineering UX.

  • Foundation embeddings

    Alim430/bioresearch-agent/skills/biomedical/foundation-embeddings Skill

    3★ repo

    Generate and evaluate single-cell foundation model embeddings (scGPT / UCE / scFoundation). Runs mock-mode pipeline validation by default — simulated embeddings that recover planted cell-type structure. Live mode stubs (pip install scgpt, UCE eval_single_anndata.py, scFoundation get_embedding.py) are documented for real deployment. Use when the user wants to embed scRNA-seq data with foundation models, compare embedding quality across models, or benchmark cell-type recovery.

  • Agent router

    Alim430/bioresearch-agent/skills/core/agent-router Skill

    3★ repo

    Classify a free-text research intent into the right BioResearch Agent workflow (literature / biomarker / causal / case-study / doctor) using a deterministic, rule-based router. Use when an AI client receives an open-ended biomedical research request and must pick which workflow or skill to invoke — NOT to generate or evaluate hypotheses. No LLM, no network.

  • Copilot

    vibhasdutta/loop-engineer/platforms/copilot Skill

    no license3★ repo

    Domain-agnostic autonomous loop wizard for VS Code GitHub Copilot. Asks 3 questions, then orchestrates a fully autonomous agent team via true parallel subagent dispatch. Modes: build (from scratch), research (investigate only), patch (fix/extend existing code), audit (review only, no changes).

  • Hermes

    vibhasdutta/loop-engineer/platforms/hermes Skill

    no license3★ repo

    Loop engineering wizard for Hermes Agent. Asks 3 questions, then orchestrates a fully autonomous parallel agent team (resource-scout, researcher, planner, agent-factory, executor, verifier, auditor, memory-keeper) for any goal. Uses delegate_task for true parallel subagent dispatch. Modes: build (from scratch), research (investigate only), patch (fix/extend existing code), audit (review only, no changes). Persistent memory, git integration. Activate with /loop-engineer.

  • Prd writing

    limengzhe27-boop/claude-product-doc-skills/skills/prd-writing Skill

    3★ repo

    完整项目规范(PRD)生成器——把一个产品想法转化成可直接喂给 Claude Code / Cursor 的 10 模块完整规范文档。自动检测并继承 BRD.md / MRD.md 的交接区数据,避免重复提问。当用户提到"帮我写个PRD"、"我想做一个XX"、"帮我梳理需求"、"我要vibe code一个产品"、"帮我写项目规范"、"coding spec"时立即触发。也适用于"帮我把这个想法变成可执行的文档"、"我想vibe code一个产品出来"、"帮我理一下做什么"等表达。即使用户只说"我想做个XX"或"帮我想想这个怎么做",只要最终目的是产出一份可供 AI 编程 Agent 执行的 PRD,都应触发此 skill。与 ai-agent-prd-writer 的区分:ai-agent-prd-writer 是工作场景的正式 PRD(给开发团队看),本 skill 是给 AI 编程 Agent 看的项目规范(用户直接拿去 vibe coding)。

  • Stack handoff

    drmahdikazempour/agent-stack/skills/stack-handoff Skill

    3★ repo

    Write or resume a continuity handoff so work survives across sessions or teammates — current task, decisions made, open threads, git state, and next steps. Use when ending a session, switching context, or picking up someone else's work. Wraps `agent-stack handoff write|resume`, which stores a single `.agent-stack/handoff.md` you can edit freely.

  • Cheat predict

    LiHongwei-cn/lihongwei-cn/global-specs/skills/cheat-on-content/skills/cheat-predict Skill

    3★ repo

    给最终稿写一份 immutable 盲预测日志。这是 cheat-on-content 整个校准循环的核心动作——预测段一旦写完不可改,由 hook 强制。**自动检测**:如目标文件已有 `## 预测` / `## 预测 v1` 段(被 cheat-shoot 调用走 v2 模式),改成 append `## 预测 v2` 而非覆盖。**打分通过 Task tool 委派给 `cheat-score-blind` sub-agent**(context-isolated channel B),主 Claude review 后落盘。触发词:"启动预测"/"start prediction"/"给这稿子打分并预测"/"写预测日志"。

  • Neat freak

    LiHongwei-cn/lihongwei-cn/global-specs/skills/neat-freak Skill

    3★ repo

    End-of-session knowledge cleanup with OCD-level rigor — reconciles project docs (CLAUDE.md, README.md, docs/) and agent memory against the code so nothing rots. 会话结束后对项目文档和记忆进行洁癖级审查与同步。MUST trigger when the user says: "sync up", "tidy up docs", "update memory", "clean up docs", "/sync", "/neat", "同步一下", "整理文档", "整理一下", "更新记忆", "梳理一下", "收尾", "这个阶段做完了", "新人能直接上手", or any phrase suggesting a dev milestone where knowledge needs reconciliation. Also trigger when the user reports stale docs, conflicting memories, or wants a clean handoff to teammates or other agents. Bare "整理" / "tidy" with prior dev context counts — do not under-trigger. Cross-platform: works on Claude Code, OpenAI Codex, OpenCode, and OpenClaw.