Explore AI Agent Skills & Claude Prompts
Discover open-source agent skills for Claude Code, Codex, ChatGPT, and any tool that uses SKILL.md.
Enter through keywords, occupations, creators, and GitHub sources to see what kinds of skills are emerging across domains.
Use the same catalog through the API
Connect 381,784 public skills to your own search, analytics, or agent workflow with the REST API.
Querying local SQLite index...
add-integration-api-tool
by sirkirbyUse this skill when adding any tool that talks to the UniFi public Integration API (X-API-Key auth, /proxy/network/integration/...). The integration API uses a different ID namespace and field schema from the V2 controller API, so these tools form their own "families" with scoped IDs. Covers when to start a new family vs join an existing one, how to scope IDs in tool descriptions, the narrow conditions under which silent ID translation is allowed (rare), how to wire auth, and how to bridge between ID spaces (explicit named bridging tools only). Applies whether you are adding a read tool, a mutation tool, or a paired family of both.
myco-api-endpoint-serializer-authoring
by sirkirbyApply this skill when adding an endpoint to apps/api/, authoring a mutation-ack serializer, wiring ManagerFactory, resolving Phase 5A routing collisions, or diagnosing the validate_manifest or test_resource_route_coverage CI gates. Covers: resource vs. action classification (409 vs. 200+envelope); Cursor + paginate() pagination in services/pagination.py; ManagerFactory concurrency scoping; apps/api → unifi-core-only dependency rule; Phase 5A patterns (capability-aware dispatch on product_kinds, dual-kind stats, wrapper-dict unwrap, path disambiguation, TOOL_ROUTE_OVERRIDES in test_resource_route_coverage.py); @register_serializer(tools={...}) decorator; seven RenderKind values; field-curation discipline; _reset_registry_for_tests() isolation; and the two CI gates. Post-Phase 6: read tools use Strawberry types (graphql-api-extension skill); serializers cover mutation acks only. Activate even without explicit serializer mention — any apps/api/ PR may need CI-gate alignment.
myco-claude-plugin-config
by sirkirbyCovers plugin configuration, transport stability, local development workflow, and release lifecycle for the three Claude MCP plugins in this repo (unifi-network, unifi-protect, unifi-access). Apply this skill when modifying plugin.json files, changing transport.py asyncio patterns, writing or auditing shell scripts in plugins/unifi-*/scripts/, verifying a local plugin load against the marketplace cache, or cutting a release for plugin-only config changes — even if the user does not explicitly ask about plugin stability or MCP transport architecture. Procedures: (1) set UNIFI_MCP_HTTP_FORCE correctly in plugin.json, (2) maintain the stdio-primary transport pattern in transport.py, (3) distinguish claude --plugin-dir from marketplace installs during local dev, (4) run check-prereqs.sh before plugin activation changes, (5) keep shell scripts Bash 3.2-compatible for macOS, and (6) issue no-op patch releases when only plugin.json changes.
myco-claude-plugin-config-transport
by sirkirbyActivate when working on Claude plugin configuration, MCP transport behavior, or plugin release mechanics in unifi-mcp. Covers: keeping UNIFI_MCP_HTTP_FORCE at its safe default in apps/*/config/config.yaml for all three apps (network, protect, access); the stdio-primary transport pattern in transport.py and why asyncio.wait(FIRST_COMPLETED) must not be reintroduced; local development with claude --plugin-dir and its marketplace-cache blind spot; running check-prereqs.sh before plugin setup; writing Bash 3.2-compatible shell scripts for plugins/; multi-target plugin support for Claude, Codex, and OpenClaw targets; atomic version sync across plugin manifests; and issuing no-op patch releases when only plugin config or scripts change. Apply even if the user doesn't explicitly mention transport races — activate whenever plugin behavior, MCP tool availability, or plugin-scoped shell scripts are being modified.
myco-community-pr-review
by sirkirbyUse this skill when reviewing or merging any community PR in unifi-mcp — even if the user just says "take a look at this PR" or "can we merge this." Covers the quality gate checklist (f-string logger ban, Ruff lint, validator registration, doc site update ordering), the fork-edit model for trusted contributors, org-fork push limitations, dual-subagent review, PR body standards, live smoke tests, mutating cycles, the unresponsive-first-time-contributor fork-edit exception, and the close-and-redirect pattern. Also covers community infrastructure setup (.github/ health files, issue routing, bug report template design) and evidence-first bug triage protocol. Apply this skill before approving any externally-authored PR, before running the merge command, when auditing recently merged PRs, and when setting up community engagement infrastructure.
myco-extend-unifi-api
by sirkirbyApply this skill when implementing a new UniFi resource type end-to-end across all layers — manager class, tool layer, domain models, tests, API REST/GraphQL exposure, and action dispatcher integration. Covers: manager CRUD with 405 workarounds, V2 API response normalization, domain Pydantic models and field validation, tool modules with preview/confirm flow, typed action input models for non-CRUD operations, test suites at both layers, manifest generation, Strawberry GraphQL type registration, cursor-based pagination for list endpoints, render-hint conventions, HTTP error contracts (409 for capability mismatch), ManagerFactory multi-controller concurrency, the multi-surface Phase 8 CI gate, field-symmetry migration procedure, update tool fetch-merge-put pattern, mutation tool registration, and DISPATCH_ARG_TRANSLATORS action dispatcher wiring. Activates for any task that introduces new resource support across the manager/tool/API boundary.
myco-live-smoke-testing
by sirkirbyActivate this skill when running, interpreting, extending, or debugging the live hardware smoke test harness in `scripts/live_smoke.py`. Covers all aspects of manifest-driven live testing against real UniFi hardware: .env credential setup, tool classification tiers, --phase flag selection to bound blast radius, the human-in-the-loop confirmation gate for mutations, artifact interpretation in live-smoke-results/, adding new tools to the harness, and recognizing known API contract failure patterns that mock-based CI cannot catch. Apply this skill even if the user doesn't explicitly ask about the harness — activate whenever a PR requires live smoke evidence, a new tool category needs coverage, or an API contract mismatch is suspected. CRITICAL: Treat live smoke as a PRE-MERGE BLOCKING GATE when code changes API response parsing (new fields, filtering logic, payload normalization).
myco-monorepo-release-pipeline
by sirkirbyCovers the full release pipeline for the unifi-mcp monorepo: determining release scope by analyzing changed packages, scoping hatch-vcs version tag globs per Python package to prevent sibling-tag contamination, pushing tags in strict dependency order (unifi-core → unifi-mcp-shared → app servers → relay → worker when needed), configuring scripts/generate_release_notes.py path scoping per package, wiring per-package publish workflows for OIDC trusted publishing, coordinating cross-package version bumps in pyproject.toml, understanding app vs. library versioning and writeback behavior, validating releases post-tag, and verifying shared package architecture constraints (DI-only rule, scope gate, relay sync) before releasing shared packages. Apply when cutting any release, bumping unifi-core, or debugging a versioning failure. CRITICAL: All PRs require pin-alignment CI gate before merge (automated, cannot be skipped).
myco-runtime-bootstrap-and-test-isolation
by sirkirbyActivate this skill when adding a new manager, adding a new tool category, writing or debugging tool unit tests, diagnosing tool-visibility failures, or investigating startup performance — even if the user doesn't explicitly ask about the runtime architecture. Covers four procedural domains: (1) the four-stage bootstrap sequence and which layer owns which changes, (2) adding a new manager via @lru_cache singleton factories in runtime.py, (3) the three-stage decorator replacement system that enables test isolation without a live controller, and (4) lazy loading mechanics and tools_manifest.json regeneration that controls which tool categories are visible to tool_index.
myco-two-tier-tool-discovery
by sirkirbyActivate this skill whenever you need to extend, maintain, or debug the two-tier tool discovery system — even if the user doesn't explicitly ask about "tool discovery." This covers: adding or modifying parameters on the tool discovery layer in `tool_index.py` (including the critical FastMCP named-param wrapper pattern vs. the silent `args: dict` trap), implementing or adjusting the compact-vs-full schema tiers (`include_schemas` flag, token-count rationale), keeping the Worker cloud path in parity with local `tool_index` changes, updating `plugins/*/skills/*/SKILL.md` runtime manifests after any parameter or behavior change, and preserving the one-unified-server architecture. Also applies to future MCP spec issue #2211 (`max_response_bytes`) alignment.
myco-unifi-client-data-correctness
by sirkirbyApply this skill whenever modifying, debugging, or extending UniFi client data retrieval in the unifi-mcp monorepo — even if the user doesn't explicitly ask about correctness or fallback behavior. Covers five permanent architectural fixtures of clients.py: (1) choosing the right endpoint (/stat/sta for live clients, /rest/user for offline or historical only); (2) preserving name and hostname as independent fields rather than collapsing with "or"; (3) deriving online status via the _is_online() uptime-field fallback when is_online is absent from the payload; (4) building a resilient fallback chain that never re-raises transient endpoint errors; (5) merging dual-source raw payloads with live data winning on overlapping keys. Verified against a live UDM SE controller (UniFi OS 5.1.12 / Network App 10.3.58): 0/180 active clients carry is_online; 121/445 clients have name and hostname set to different values.
incident-investigation
by sirkirbyInvestigate a network incident by correlating device events with camera footage and physical access logs. Use when the user reports a device going offline, a network anomaly, or wants to understand what caused an infrastructure event.
Browse Agent Skills by Occupation
23 major groups · 867 SOC occupations
Browse by Category
Explore agent skills organized by their primary use case
Explore the agent skills ecosystem by occupation and creator
SkillMD is not just a keyword search box. It is an open map that organizes public skills by occupation, creator, and repository, helping you see which workflows, judgment criteria, and domain habits people are writing for AI agents.
Then follow creators and GitHub repositories back to the source: compare the skills a team maintains, whether the repo is active, and how the README frames the work before you open, install, or reuse anything.
Use it three ways: learn an unfamiliar field by occupation, study how creators organize skills, then use source context to decide what is worth opening or reusing.
01 Map a field
Browse 23 occupation groups and 867 SOC roles to learn what skills exist in adjacent domains and how they break down real work.
02 Follow creators
Use creator and repository pages to inspect maintained skill collections, recent updates, and source context before trusting a result.
03 Search with sources
Search 1.7M+ collected skills, then use occupation tags, creators, and GitHub source context to decide what is worth opening.
Start with the occupation map, then follow creators and repositories back to real code. SkillMD helps explain why a skill is worth opening, not only what it is named.
Standardizing Agent Capabilities with SKILL.md and Model Context Protocol (MCP)
In the rapidly evolving landscape of artificial intelligence, LLM agents (Large Language Model agents) have transitioned from simple text predictors to autonomous problem solvers. To orchestrate complex, multi-step agentic workflows, developers require a standardized format to specify agent capabilities, prompt instructions, system rules, and database bindings. This is where SKILL.md and the Model Context Protocol (MCP) have emerged as standard developer paradigms. SkillMD serves as the central directory for indexing, exploring, and sharing these critical agent configurations.
Our open-source registry currently tracks over 1.7 million collected SKILL.md configurations and system prompts. By compiling agent configurations from active developers on GitHub, we bridge the gap between prompt engineering research and production execution. Whether you are building agents with Anthropic's Claude Code, OpenAI's GPT-4, Google's Gemini, or local models using Ollama and LlamaIndex, standardized skill definitions ensure your agents behave predictably across different runtime environments.
What is the Model Context Protocol (MCP)?
The Model Context Protocol (MCP) is an open-source standard designed to connect LLMs to data sources, developer tools, and external environments. MCP establishes a bidirectional communication channel between client applications (like Cursor, Claude Desktop, or custom agent systems) and servers hosting data or capabilities. Standardizing instructions via SKILL.md enables LLMs to query databases, read local files, execute terminal commands, and integrate third-party APIs. SkillMD allows you to find ready-to-run MCP servers and prompt instructions for various occupations and technical tasks.
The Structure of a Professional SKILL.md File
A valid SKILL.md configuration is designed to be easily read by humans and parsed by LLMs. It contains precise system instructions, trigger conditions, required parameters, and execution examples. Below is the typical architectural blueprint of a professional agent skill:
- Metadata & Core Scope: Declares the name of the skill, author details, target models, and a description of the capability.
- Triggers & Intent Detection: Details semantic triggers that help the agent decide when to invoke this skill.
- System Prompts: Explicit system-level instructions that direct the agent's behavior, personality, safety guardrails, and formatting preferences.
- Capabilities & Tools: Lists the files, databases, or APIs the agent must access to complete the tasks.
- Few-Shot Examples: Demonstrates real inputs and outputs, helping the model generalize behavior through in-context learning.
Optimizing Agent Workflows for Modern LLMs
Writing effective agent skills requires deep knowledge of prompt engineering. With the release of advanced reasoning models like Claude 3.5 Sonnet, ChatGPT o1, and DeepSeek-V3, prompt templates must focus on structured thinking. Developers are encouraged to use XML tags (e.g., <thought>, <context>, and <rules>) to isolate execution boundaries. Standardized prompts prevent agents from suffering from context drift, ensuring that long-running tasks remain aligned with the initial system parameters.
Exploring by SOC Occupations and Creator Profiles
What makes SkillMD unique is its taxonomy. Instead of simple text search, we parse and organize files according to the Standard Occupational Classification (SOC) system. This means you can discover skills written for Computer and Mathematical roles, Business and Financial operations, Legal, Design, and and Educational Instruction fields. By tracking creator profiles, developers can study how different teams organize their custom instructions, compare version updates, and fork public configs for specialized enterprise use cases.
SkillMD operates as a high-performance index running on a fast Go backend and a highly responsive Astro SSR frontend. All search queries execute in milliseconds, featuring smart debouncing to prevent multiple API requests while keeping user data secure. Join our community of developers to standardize your AI agent instructions and optimize your LLM prompting workflows today.
Frequently Asked Questions
A practical guide to agent skills: what they are, how to inspect them, and how SkillMD helps you explore the ecosystem.