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
b-mendoza/agent-skills/skills/generate-handoff-document Skill
no license3★ repoA collection of reusable skills for AI-assisted development workflows
b-mendoza/agent-skills/skills/improving-skill-definition Skill
no license3★ repoA collection of reusable skills for AI-assisted development workflows
i-stack/ai-coding-kit/skills-engineering/cross-model-review Skill
3★ repo跨模型对抗审查已锁定的 PLAN.md——自动发现可用 CLI(codex/gemini/claude),推荐两个不同 provider 的组合并让用户选择,reviewer 只读运行输出 VERDICT:APPROVED|REVISE,主 agent 仲裁并把理由写进 PLAN-REVIEW-LOG.md,MAX_ROUNDS 不收敛输出 deadlock。基于 chaseai-yt/grill-me-codex(MIT)的 Act 2 思路。
ddnetters/homelab-agent-skills/slite-knowledge-base Skill
3★ repoSlite knowledge base API — ask questions, search notes, retrieve content, manage users and groups, and audit knowledge health via the REST API
Adit-Jain-srm/skill-forge/skills/dynamic-workflow Skill
3★ repoDesign and orchestrate Claude Code dynamic workflows — the new multi-agent system that fans out 10-100 subagents in parallel with verification. Teaches when to use workflows vs single-agent, how to structure fan-out/fan-in patterns, and how to build verification gates. Use when tackling complex tasks that benefit from parallel research, multi-perspective analysis, or divide-and-conquer strategies. Also use when user says workflow, fan out, parallel agents, deep research, multi-agent, or wants to break a big task into verified parallel pieces.
The same file is in 1 other repository
Adit-Jain-srm/skill-forge/skills/handoff Skill
3★ repoCompact the current conversation into a handoff document so another agent or future session can continue the work without losing context. Use when ending a session, switching tasks, the context is getting long, or user says handoff, wrap up, save progress, or "continue this later".
Adit-Jain-srm/skill-forge/skills/mcp-conductor Skill
3★ repoOrchestrate multiple MCP servers together for complex multi-step tasks. Teaches agents to chain Exa search → Bright Data scraping → GitHub API → file operations in intelligent workflows. Use when a task requires data from multiple sources, when combining MCP tools for research, when building multi-step automations across services, or when the user says orchestrate, combine MCPs, multi-source, research pipeline, or chain tools together.
The same file is in 1 other repository
Adit-Jain-srm/skill-forge/skills/session-guard Skill
3★ repoPrevents long-session corruption AND context compaction amnesia. Monitors session health, detects instruction drift, enforces critical-rule recitation, and splits proactively before degradation hits. Use when working on complex multi-step tasks, when a session is getting long, when the agent starts ignoring rules it followed earlier, when CLAUDE.md conventions drift, when output quality seems to degrade, or after any context compaction event.
The same file is in 1 other repository
product-on-purpose/agent-skills-toolkit/skills/askit-build-hook Skill
3★ repoCreates and improves event-driven hooks (Advanced tier) for a plugin to the Advanced Skill Library Standard. Use when you need to add a hook, enforce a guard on a tool or session event, inject context, or document a hook's event, scope, and failure behavior.
product-on-purpose/agent-skills-toolkit/skills/askit-build-mcp Skill
3★ repoCreates and improves MCP server definitions (a portable .mcp.json) for a plugin to the Advanced Skill Library Standard. Use when you need to add an MCP server to a plugin, author or extend .mcp.json, or wire the per-target mcpServers manifest pointer.
product-on-purpose/agent-skills-toolkit/skills/askit-build-subagent Skill
3★ repoToolkit and Standard for building, grading, and scaling cross-agent skill libraries (Claude Code + Codex) to a tiered Bronze/Silver/Gold quality bar.
product-on-purpose/agent-skills-toolkit/skills/askit-capability-advisor Skill
3★ repoReports which component types a target agent can run and recommends a conformance tier before a plugin is built, mapping Claude Code and Codex capabilities to the Advanced Skill Library Standard's component types. Use when choosing agent-targets, checking whether a component is portable across agents, or deciding which tier to aim for.
product-on-purpose/agent-skills-toolkit/skills/askit-migrate Skill
3★ repoAssesses an existing skills repository against the Advanced Skill Library Standard, produces a staged bring-to-conformance plan, and writes the minimal canonical manifest so the repo becomes gradeable. Use when adopting a foreign or ad-hoc skills repo, migrating a Claude-only plugin toward cross-agent conformance, or planning a Bronze-to-Silver upgrade path.
seaartpublic/skills/prompt-preflight-advice Skill
3★ repoProvide concise, goal-aligned suggestions before generating or optimizing image, video, audio, GPT Image, or storyboard prompts. Use when the user asks for advice, direction, risks, alternatives, or "how should this prompt be improved" before producing the final prompt. This skill gives suggestions only and should not replace the main prompt-building skill.
seaartpublic/skills/storyboard-prompt-assistant Skill
3★ repoConvert concepts, scripts, ad copy, product ideas, rough stories, or vague video ideas into high-quality storyboard prompts, shot lists, and AI-video-ready multi-shot prompt tables. Use for storyboard, shot script, camera planning, shot list, video storyboard, trailer planning, script-to-storyboard, AI video prompts, and incremental storyboard edits.
DobroslavRadosavljevic/skills/skills/handoff Skill
3★ repoProduce or consume agent-to-agent handoff context so another session can resume work without re-deriving decisions, state, and next steps. Use when the user invokes $handoff, says "handoff", "hand off", "session handoff", "write a handoff", "resume from handoff", "context for the next agent", or asks to package current work for another AI agent or later session.
DobroslavRadosavljevic/skills/skills/subagents Skill
3★ repoOrchestrator/subagent coordination workflow for tasks that can be split into clear disjoint work. Use when the user invokes $subagents, says "use subagents", "spawn agents", "parallelize this", "delegate this", "split this across agents", or asks for coordinated multi-agent work. Prefer subagents for harder tasks, broad research, multi-file changes, independent reviews, or separable implementation lanes. Skip subagents when the task is simple, the harness does not support them, or safe disjoint work cannot be identified.
oubakiou/delegate-skills/skills/delegate-imagegen Skill
3★ repo画像生成 capability を持たない requester から、画像生成 capability を持つ Codex 子プロセスへ画像生成・画像編集作業を委譲するスキル。 ユーザーが画像、イラスト、写真風ビジュアル、バナー、サムネイル、アイコン案、テクスチャ、モック画像などの生成や編集を求め、 main agent 側で画像生成手段を直接使えない、またはプロンプト試行錯誤・生成パラメータ・失敗ログを隔離したい場合に使う。 主目的は token cost 削減ではなく capability bridge と context isolation。DELEGATE_IMAGEGEN_MODEL で Codex モデルを切り替える。 imagegen の作業を委譲する場合は、この skill を使う。generic な subagent で代替しない。
oubakiou/delegate-skills/skills/delegate-x-research Skill
3★ repox.com / X 上の投稿・アカウント・話題・反応を調査する作業を、X 調査 capability を持つ子プロセスへ委譲するスキル。 ユーザーが X、x.com、Twitter、ポスト、スレッド、アカウント、ハッシュタグ、炎上、世論、直近の反応、 コミュニティノート、投稿URLの確認などを調べたい場合に使う。main agent 側で X を直接確認できない、 または最新投稿・検索結果・引用元確認を隔離したい場合に使う。現在のバックエンドは Grok CLI で、 DELEGATE_X_RESEARCH_MODEL でモデルを切り替える。 x.com 調査を委譲する場合は、この skill を使う。generic な subagent で代替しない。
goshippo/ai/providers/claude/plugin/skills/upgrade-shippo Skill
3★ repoGuide for Shippo API version changes, webhook payload versioning, and how the hosted MCP server handles updates. Use when reasoning about backward compatibility, handling new fields in webhook payloads or API responses, troubleshooting OAuth/version-mismatch errors against the hosted MCP, or auditing an existing Shippo integration before a change.
goshippo/ai/providers/codex/plugin/skills/upgrade-shippo Skill
3★ repoGuide for Shippo API version changes, webhook payload versioning, and how the hosted MCP server handles updates. Use when reasoning about backward compatibility, handling new fields in webhook payloads or API responses, troubleshooting OAuth/version-mismatch errors against the hosted MCP, or auditing an existing Shippo integration before a change.
goshippo/ai/skills/upgrade-shippo Skill
3★ repoGuide for Shippo API version changes, webhook payload versioning, and how the hosted MCP server handles updates. Use when reasoning about backward compatibility, handling new fields in webhook payloads or API responses, troubleshooting OAuth/version-mismatch errors against the hosted MCP, or auditing an existing Shippo integration before a change.
davidlee/doctrine/plugins/doctrine/skills/dispatch-agent Skill
3★ repoThe claude arm of `/dispatch` — spawn a worker via the `Agent` tool using the dispatch-worker subagent type with worktree isolation. Base is explicit — `dispatch arm-spawn --base B` writes the base file, then cd into the spawn dir before the Agent spawn so the WorktreeCreate hook forks at B. Worker self-commits via the gated `worker_commit` MCP tool; the orchestrator then does whatever `doctrine dispatch next` prescribes, one action at a time. Reached only from the `/dispatch` router on a claude↔env-marker agreement; do not invoke directly.
davidlee/doctrine/plugins/doctrine/skills/dispatch Skill
3★ repoUse to drive a slice's phases to completion through sub-agent workers in isolated worktrees — you orchestrate and are the sole writer, the workers execute. Routes to `/dispatch-subprocess` (codex/pi) or `/dispatch-agent` (claude); overridable via `[dispatch] claude-force-subprocess-dispatch` in `doctrine.toml`. The per-phase funnel is driven by `doctrine dispatch next`, which prescribes exactly one action at a time and is identical on both arms. Default serial (one worker per phase); parallelize file-disjoint phases. Conflicts report-and-halt, never auto-merge.
davidlee/doctrine/plugins/doctrine/skills/record-memory Skill
3★ repoUse when you discover or confirm a durable fact, constraint, invariant, convention, sharp edge, footgun, or reusable workflow that a future agent would otherwise rediscover — capture it before it is lost to conversation context. Triggers at phase/slice wrap-up too.
davidlee/doctrine/plugins/doctrine/skills/retrieve-memory Skill
3★ repoUse before making non-trivial assumptions — before touching a subsystem you have not this session, before running or changing a command pipeline, when code and docs conflict, when asked "what is the right way here?", when debugging a recurring failure, or when about to answer with "probably/usually/likely".
davidlee/doctrine/plugins/doctrine/skills/reviewing-memory Skill
3★ repoStructured audit for stability gates — before releases, migrations, large refactors, or when agent confusion is detected. Use to prune, correct, and re-anchor the memory corpus before it drifts past useful.
davidlee/doctrine/plugins/doctrine/skills/worktree Skill
3★ repoUse when work needs isolation from the current workspace — when /execute (mode=solo) or the /dispatch funnel must run a phase on its own branch without touching the main working tree. Detects existing isolation, creates a fork via `doctrine worktree fork` (the sole creation+provision+mark verb; coordination tier excluded), runs the commit-before-spawn / branch-point / baseline guards, and hands back the fork branch.
download4you/n2-fieldops/fieldops-engineering Skill
3★ repoDiagnoses, implements, and verifies non-trivial or risk-sensitive changes in an existing repository while preserving the user's uncommitted work, following repository instructions (AGENTS.md or CLAUDE.md), keeping edits within authorized scope, and handing off with runtime evidence. Use when the user asks to fix a bug, add or change a feature, refactor, migrate, edit configuration, chase down a regression, or make a change safely amid overlapping edits or worktrees where verification and change control matter. Not for isolated code snippets or simple explanations; for diagnosis-only requests, prove and explain the cause without fixing unless asked.
download4you/n2-fieldops/fieldops-orchestrator Skill
3★ repoCoordinates multiple FieldOps skills for complex, unfamiliar, or mixed-domain work spanning research, prompt control, authorized CTF evidence, repository changes, PowerShell Unicode handling, or prompt refinement. Use when a task needs two or more FieldOps workflows, when the correct specialist is unclear, or when the user invokes fieldops-orchestrator. Prefer a narrower skill for clearly single-domain tasks.
download4you/n2-fieldops/fieldops-prompt-decorators Skill
3★ repoInterprets and executes Prompt Decorators written with +++Name(key=value) syntax, handling message vs. chat scope, parameter validation, decorator composition, conflict resolution, retained-state inspection (ActiveDecs and AvailableDecs), selective clearing, and conversation export or dump. Use when a message contains a supported +++ token (for example +++Planning, +++StepByStep, +++Reasoning, +++Interactive, +++Refine, +++Export, +++Dump, +++ChatScope, or +++MessageScope) or when the user asks to create, validate, debug, or explain decorator behavior. Do not trigger for ordinary tone, plan, or formatting requests that contain no +++ decorator syntax or discussion.
download4you/n2-fieldops/fieldops-prompt-refiner Skill
3★ repoTransforms rough, ambiguous, overloaded, or reusable requests into precise, execution-ready prompts for coding agents. Use when asked to improve, rewrite, refine, tighten, structure, modularize, audit, or professionalize a prompt, AGENTS.md instruction set, task brief, system prompt, or agent workflow. Preserves intent while clarifying outcome, scope, inputs, constraints, authority, verification, and deliverables. Does not apply when merely executing an already well-specified task.
fxckcode/chatwoot-skills/skills/chatwoot-captain Skill
3★ repoSet up and use Captain AI — Chatwoot's built-in AI agent for automated customer support. Use when configuring AI assistants, FAQ bots, copilot features, or custom tools for Captain.
ulpi-io/autonomous-engineering/auto-learn Skill
no license3★ repoClose every autonomous run with a learning pass so the machine never repeats a mistake it already paid for: harvest the run's STRUCTURED artifacts (checkpoint register, blocked-task reasons, guard blocks, fix-loop counts, degradations), verify each candidate lesson adversarially, dedupe against what is already known, then WRITE each learning to the Claude Code memory that LOADS AUTOMATICALLY next session — a convention or plan-shape rule into CLAUDE.md, an area-specific lesson into a path-scoped .claude/rules file, an environment quirk into auto memory. Machine defects are surfaced to the user, never self-patched. Because those are native context files Claude Code already loads, the next run picks the lessons up with no extra step. Use as the closing phase of every run, or standalone.
ulpi-io/autonomous-engineering/auto-map Skill
no license3★ repoGenerate and refresh the project's CONTEXT ARCHITECTURE — a disclosure-tiered map so the agent starts every session knowing the repo without paying for what it doesn't need. Platform-aware by explicit contract (claude|codex|dual): for Claude a lean root CLAUDE.md (loads always), path-scoped .claude/rules/*.md (load only when touching matching files), nested CLAUDE.md throughout the tree (loaded on demand — zero cost until then); for Codex the root + nested AGENTS.md tree instead — never inventing a memory location the platform lacks, never touching private agent memory. Deep references are LINKED, never imported. Every claim is VERIFIED against the real repo before it ships (commands actually run; paths actually exist) — the map cannot lie. Use to initialize project context, refresh it after meaningful changes, or as the pipeline's final phase so the map reflects what just shipped.
ulpi-io/autonomous-engineering/schedule-recurring-agent Skill
no license3★ repoStand up a recurring scheduled agent for standing work (triage, monitoring, audits, digests): a self-contained IDEMPOTENT brief (each run wakes memory-less and must dedup prior work), a cadence matched to how often work actually arrives, a per-run BOUND, escalation rules, and a teardown condition. For durable unattended work that runs while you're offline, use claude.ai Routines (the /schedule skill / RemoteTrigger); an in-session CronCreate is the lighter, session-scoped alternative. Use for scheduled repeat work — not one-off waits (watch-and-act).
Xuan0629/shared-skills/handoff Skill
no license3★ repoOne source. Every agent. Zero drift. | 跨 Agent 共享 Skill 管理系统 — 单一数据源,同步到 Claude Code / Codex / Hermes / OpenClaw
BrianV1981/aim-skill-library/skills/aim-memory-search Skill
3★ repoSearch A.I.M. Engram / hybrid memory via the vessel ./aim CLI. Use for project history and factual recall. Slash: /aim-memory-search.
urmzd/dotfiles/dot_agents/skills/configure-ai Skill
3★ repoAuthors AGENTS.md and defines the skills-vs-docs boundary for a repo. Configures AI coding tools (Claude Code, Codex, Antigravity CLI, Copilot) and the skills-as-docs philosophy. Use when creating or auditing AGENTS.md, setting up AI tooling for a project, or deciding what belongs in a skill versus docs/. Do NOT use for README structure -> use write-readme; for llms.txt generation -> use create-llms-txt; for Claude-specific skill features (invocation control, subagents, model overrides) -> use extend-oss-skills-to-claude.
urmzd/dotfiles/dot_agents/skills/extend-oss-skills-to-claude Skill
3★ repoExtend standard agentskills.io skills with Claude Code-specific features. Invocation control, subagent execution, dynamic context injection, string substitutions, model/effort overrides, and deployment scoping. Use when adapting a portable skill for Claude Code, adding Claude-specific frontmatter, setting up subagent delegation, or configuring skill permissions.
urmzd/dotfiles/dot_agents/skills/orchestrate-agents Skill
3★ repoOrchestrate multiple agent CLIs (Claude, Codex, Antigravity) via tmux with a shared fleet store, dispatching one guardian subagent per pane. Survey-first: inspects and adopts existing tmux sessions, windows, and agent panes before creating anything new. Use when running a multi-agent session, dispatching parallel work, or coordinating subagents across panes.
urmzd/dotfiles/dot_agents/skills/run-eval-harness Skill
3★ repoRun a project's own eval suite end to end without babysitting: locate the harness, launch it in the background, monitor to completion, parse the report into a normalized summary (pass rate, latency percentiles, tokens, cost), and diff against the previous run with per-case regressions explained. Use when asked to "run the evals", re-run after a change, or score a new dataset in any repo with an eval harness. Do NOT use for the GAP benchmark suite (use run-evals), SAIGE provider-boundary checks (use verify), or designing new eval sets (that is authoring work, not a run).
Bosh-Kuo/awesome-agent-toolkit/examples/.agents/skills/skill-creator-anthropic Skill
no license3★ repoGuide 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.
Bosh-Kuo/awesome-agent-toolkit/examples/.agents/skills/skill-creator-openai Skill
no license3★ repoGuide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
Bosh-Kuo/awesome-agent-toolkit/examples/.claude/skills/skill-creator-anthropic Skill
no license3★ repoGuide 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.
Bosh-Kuo/awesome-agent-toolkit/examples/.claude/skills/skill-creator-openai Skill
no license3★ repoGuide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
Bosh-Kuo/awesome-agent-toolkit/skills/skill-creator-anthropic Skill
no license3★ repoGuide 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.
Bosh-Kuo/awesome-agent-toolkit/skills/skill-creator-openai Skill
no license3★ repoGuide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
agent-blueprint/agent-blueprint-skills/plugins/agent-blueprint-skills/skills/research-agent Skill
no license3★ repoMulti-tool research methodology with automatic depth routing for company research, market analysis, deep research, and competitive investigation. Routes queries through free tools first (web search, page scraping), escalates to paid tools only when needed (Perplexity, Gemini Deep Research). Includes caching, source verification, and structured output. Use when researching companies, topics, markets, competitors, or any question requiring web information. Triggers: research, look into, find out about, investigate, what do we know about, deep research, company research, market analysis.
agent-blueprint/agent-blueprint-skills/plugins/agent-blueprint-skills/skills/servicenow-ai-agents Skill
no license3★ repoServiceNow AI Agent Studio reference for coding agents deploying AI agents on ServiceNow: sn_aia data model, agentic workflows, and Now Assist agents. Covers entity mapping, URL patterns, tool types, prompting patterns, MCP setup, and browser navigation. Use when building, configuring, or debugging AI agents on ServiceNow (Zurich or later). Triggers: servicenow, AI Agent Studio, agent studio, sn_aia, agentic workflow, now assist agents, servicenow agent deployment.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/agent-design/knowledge-injection-strategy Skill
no license3★ repoChoose between RAG, fine-tuning, and context engineering for domain knowledge. Use when deciding how your agent will 'know' things, when responses feel generic despite good prompting, or scaling from MVP to production.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/agent-design/system-prompt-engineering Skill
no license3★ repoWrite a production-grade system prompt using the five-layer SOUL structure. Use when building a new agent SOUL, debugging inconsistent agent identity under pressure, or refactoring a prompt that has grown unmaintainable.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/build-skills/agent-eval-framework Skill
no license3★ repoBuild a three-layer eval system: vibe check, golden dataset, production signals. Use when your agent 'seems to work' but you can't quantify it, before deciding to ship, or diagnosing why users are churning.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/build-skills/multi-agent-orchestration Skill
no license3★ repoDesign router/worker multi-agent architecture with typed handoff contracts and failure isolation. Use when splitting a single agent, designing agent-to-agent communication, or debugging cascading failures in a multi-agent system.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/mental-models/agent-loop-model Skill
no license3★ repoExplain the perceive-reason-act-observe loop that every agent runs. Use when designing agent architecture for the first time, debugging unexpected agent behavior, or explaining 'what is an agent' to a stakeholder.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/mental-models/llm-capability-boundaries Skill
no license3★ repoMap what LLMs do reliably, unreliably, and cannot do. Use before scoping agent capabilities, when debugging surprising model behavior, or deciding whether a task needs a tool vs. the model alone.
MedocMay/ai-native-builder-consultant-skills/ai-native-builder-consultant-skills/skills/mental-models/uncertainty-and-hallucination Skill
no license3★ repoDiagnose why an agent produces confident wrong output and apply the right fix. Use when an agent is hallucinating, when designing a fact-sensitive feature, or auditing a product's risk profile.
jovd83/restassured-skill/documentation/session-state Skill
3★ repoUse when Codex needs to persist a durable resume checkpoint for Rest Assured API testing work so a human or later agent can see what is done, what is blocked, and exactly how to continue.
jovd83/restassured-skill/orchestrator Skill
3★ repoUse when Codex receives a generic Rest Assured or service-testing request and must route the work into planning, implementation, documentation, execution, CI, virtualization, or test-management flows.
MarecGents/marec-agent-skills/skills/skill-install-manager Skill
3★ repo技能安装管理器 — 自动读取技能列表文件,与当前全局已安装技能进行对比, 找出未安装或存在更新的技能,并执行一键安装/更新。支持三级回退: HTTPS → SSH → MCP/GitHub API 直接抓取。使用 `npx skills` 命令安装, 安装为 Global,指定 Reasonix / Claude Code / OpenCode 三个 Agent。 当用户提到"检查技能"、"同步技能"、"安装技能"、"技能管理"、 "skill update"、"skill sync"、"批量安装"、"缺少技能"等任何与 技能安装管理相关的需求时,务必使用本技能。