Muse Skill Hub

An open, LLM-agnostic capability runtime for AI agents. Define capabilities once. Discover them dynamically. Execute them through typed tools. Enforce permissions at runtime. Keep failures honest. Created by Wahyu Nur Iman.

⚡ Executable Runtime

Generated from registry + test suite — do not hand-edit. Run python tools/sync_readme.py.

Version 2.3.0 · 97 skills · 96 executable drivers (1 honest stubs) · 222 MCP tools · 309 passing tests

Evidence honesty: of the 96 executable drivers, 2 are live-tested against real provider servers (github, podcast), 3 have mock-contract tests (stripe, lovable, replit), and 91 are structural (real code, provider handshake is the operator's step). Full per-driver matrix: certification/PROVIDER_MATRIX.md (generated — do not hand-edit).

This repo is no longer just a catalog — it ships a real MCP server
exposing every action as its own typed tool (e.g. github_search_repositories,
with full JSON schemas — no opaque params blob). Skills are backed by real,
executable drivers; a few remain honest catalog-only stubs (structured
driver_not_implemented — never fake data):

cd runtime && pip install -r requirements.txt && python -m skillhub.server

94 real drivers across three kinds:
- Third-party APIs (62): github, slack, stripe, shopify, linear, vercel, asana, notion, todoist, places-search, zoom, gmail, google-calendar, google-sheets, google-drive, spotify, instagram, meta-threads, threads, facebook, dropbox, ticketmaster, image-search, flightaware, box, calendly, canva, duffel, figma, ghl, klaviyo, meta-ads, plaid (sandbox), printify, quickbooks, tts, voice-design, voice-selector, voice-calls, zapier, outlook-calendar, outlook-mail, outlook-contacts, google-contacts, google-docs, google-forms, google-slides, google-tasks, messenger, instagram-messages, threads-messages, withings, tailscale, tessie, peloton, philips-hue, podcast, shopping, wide-research, social-content-performance, evernote, healthex (MCP passthrough), granola, generate_podcast (needs ElevenLabs key for TTS)
- Local reference implementations (32): secure-vault (encrypted), permission-model, personal-feed, idea-management, goals, share-ideas, agent-library, connector-management, paired-devices, data-control, messaging-channels, wallet, apple-healthkit + google-health-connect (local export readers), device-data, media-library, forget, self-awareness, skill-creator, function-health, travel-planning, muse_db (SQLite), muse-feedback, subscription-status, wearable-device-skills, wearables-comms, booking (router), opentable (deep links), facebook-cli (passthrough), magic-moment (local ffmpeg)
- Catalog-only stubs (3): lovable, muse-early-access, replit — no public API exists (or it is an internal-only program); documented honestly instead of faked, with workarounds where one exists
MCP tools (per-action, fully typed) + skillhub_search_capabilities for dynamic capability discovery (counts in the metrics block above)
Approval engine — write actions need approval; sensitive/destructive/communication/financial/account/device risks require a real approval_id (bound to exact skill/action/params, single-use, 10-min TTL). Bare confirm=true only suffices for plain write risk
Policy engine — risk-based allow/approval_required/deny with per-action classification and custom policy files
Audit log — every execution recorded (request_id, actor, risk, params hash, approval_id, duration); secrets never logged raw; query via function-health query_audit_log
Idempotency — idempotency_key on write actions: repeats return the first result instead of re-executing
Credential manager — env → local file → encrypted vault; ref:vault:<name> references resolve server-side so secret values never reach the LLM
Hardened HTTP layer — persistent connection pooling, retry with exponential backoff + Retry-After, X-Request-ID correlation, structured 429/5xx mapping
Skill manifests — runtime/skillhub/catalog/<name>/manifest.yaml is the machine-readable contract; skillhub validate runs the conformance test (schemas, risk, auth, manifest/driver/SKILL.md drift, secret scan)
Passing tests (count in the metrics block above), including live api.github.com calls, an end-to-end MCP stdio session with per-action tools, approval-lifecycle/idempotency/audit/security tests, and credential-error coverage for every API driver

What's new in v2.2.0 (concurrency + crash hardening)

Cross-platform file locking — new skillhub.filelock module (fcntl on Unix, msvcrt on Windows); no more top-level import fcntl, so the runtime imports cleanly on Windows
Crash-atomic stores — every read-modify-write cycle (locked_json, credential store) now holds a separate <name>.lock file and writes back via tmp + fsync + os.replace: atomic against concurrency and crashes
OAuth single-flight refresh — the check → refresh → persist cycle runs under the store lock with double-checked expiry; 20 concurrent workers on an expired token produce exactly 1 refresh request
Fail-closed credential store — a corrupt/undecryptable credentials.enc raises CredentialStoreCorruptError (with a .corrupt backup) instead of silently looking like "no credentials"
Approval privacy — the approval store no longer persists raw params; only params_hash + a redacted preview. Execution re-supplies params and consume() verifies them against the hash
Real JSON Schema — validation now runs on the standard jsonschema library (Draft 2020-12, full vocabulary: $ref, if/then/else, dependentRequired, …) with format checking; skillhub validate lints every action schema
Declared dependencies — cryptography and jsonschema are now official dependencies (they were already required at runtime)
Generated README metrics — runtime/tools/sync_readme.py regenerates the metrics block from the registry + test suite; a test fails if the README drifts
Multi-process race tests — approval consumption and idempotency reservation are now proven with 20 processes (not just threads), matching the documented guarantee

What's new in v2.1.0 (security hardening)

Atomic approval consumption — approval approve/consume transitions happen inside a locked critical section, so concurrent workers can't double-spend one approval (proven by a 20-thread race test)
Atomic idempotency — PENDING → SUCCEEDED | FAILED → CONFLICT reservation happens before side effects; concurrent duplicates see pending and never re-execute; failed records allow exactly one reclaim
Honest naming — idempotent renamed to supports_idempotency_key: the runtime supports dedup keys, it does not claim mathematical idempotence of the underlying API
Output contracts — handler results are validated against each action's output_schema before success is committed; violations never get cached
Real JSON-Schema validation (v2.2.0: standard jsonschema library, Draft 2020-12) — strict by default (unknown params rejected, like MCP's additionalProperties: false)
Credential scopes — actions declare required_scopes (Gmail/Outlook/Google Workspace drivers); mismatches are blocked, unknown scopes are allowed but marked unverified in audit
Encrypted credential store — local credentials now live in an encrypted credentials.enc (Fernet, secure-vault key); the old plaintext fallback is gone, legacy files are migrated and deleted
OAuth refresh manager — encrypted access/refresh tokens with auto-refresh before expiry and rotation persistence
Audit hash chain — every event carries prev_hash/event_hash; python -m skillhub.cli audit-verify detects tampering (tamper-evident, not tamper-proof)
TF-IDF capability discovery — skillhub_search_capabilities now ranks by lexical relevance instead of substring matching
Provider mock harness — SKILLHUB_MOCK=1 routes api_request to canned responders; mock contract tests for GitHub and Stripe assert request shapes and auth flow without network
SQL hardening — muse_db strips comments/strings before keyword scanning, runs reads on a read-only SQLite connection, and enforces single statements
Zero doc warnings — all 178 missing action docs generated; skillhub validate passes clean
Passing tests (count in the metrics block above; was 67 in v1.5.0), including 20-thread concurrency races, scope mismatches, encrypted-store migration, OAuth refresh, audit tampering, and provider contracts

Security model (honest): this is a trusted, local, single-user MCP runtime. Encryption protects against casual disk exposure, not against malware or same-user attackers. actor is a local session label, not authentication. Multi-user tenant isolation, distributed exactly-once, and workflow/saga engines remain out of scope (see specs/).

See runtime/README.md for setup, driver docs, and how to add your own driver. Specs live in specs/ (manifest, error contract, approval protocol).

About this repo

This repo contains a catalog of all of Muse's skills/capabilities (Muse is Meta's personal AI agent).
The goal is an open reference: what Muse can do, what the patterns are, and inspiration for building similar capabilities in other AIs/LLMs.

Important note: This file only contains high-level lists and descriptions, not raw internal files. Muse's real skills run on a specialized runtime (MCP servers, OAuth, custom CLIs, etc.), so just reading this list won't automatically make another AI 99.9% like Muse — but it can be a very useful blueprint.

Total documented skills: 97

Universal Skills (Usable by Any LLM) 🌐

Each skill now has two files:

runtime/skillhub/catalog/<name>/README.md — Human-readable documentation (what it does, when to use)
runtime/skillhub/catalog/<name>/SKILL.md — Universal skill definition in open format (frontmatter + instructions) that any LLM can load

See USAGE.md for how to use these with GPT, Claude, Gemini, Llama, LangChain, etc.

Quick start for any LLM:

# Option 1: Paste into system prompt
with open('runtime/skillhub/catalog/github/SKILL.md') as f:
    skill = f.read()
# Add to your LLM's context

# Option 2: LangChain
from langchain.tools import Tool
tool = Tool(name="github", description="...", func=your_impl)

Skill List

Skill	Title	Description
`agent-library`	Agent Library	Manage the user's file library: uploads, generated artifacts, expiring public share links, and the media collection.
`apple-healthkit`	Apple Health	The user's synced Apple Health (HealthKit) data: daily metrics (steps, distance, calories, heart rate, HRV, VO2max), sleep sessions (stages, quality, efficiency), and workouts.
`asana`	asana	Manage Asana work: create and list tasks, assign owners, set due dates, organize projects, and track team progress.
`booking`	booking	Primary entry point for direct flight, hotel, restaurant, or event-ticket transactions and for bounded live availability checks delegated by Travel Planning. Always use before provider-specific skills
`box`	box	Search, read, upload, download, move, rename, delete, restore, and share Box content; manage comments and metadata.
`calendly`	calendly	View Calendly events and event types, and manage scheduling data using the Calendly CLI.
`canva`	canva	Design with Canva: create and edit designs and presentations, manage brand assets and folders, and generate share links.
`connector-management`	Connector Management	Manage third-party service connectors: check connection status and scopes, guide OAuth connect flows, and disconnect services.
`data-control`	Data Control	Handle user data rights: explain data collection and use, export chats and files, delete data on request, manage training opt-outs.
`device-data`	Device Data	Read cached contacts and calendar events from Muse storage. Delete Muse's local copy of either source without modifying paired devices.
`dropbox`	dropbox	Manage Dropbox storage: upload and download files, create share links, organize folders, and check space usage.
`duffel`	duffel	Use Duffel to search, book, pay for, or manage flights. Use Duffel to monitor an already booked flight's fare when the user directly asks for ongoing price monitoring.
`evernote`	evernote	Read and create notes through Evernote's official MCP server.
`facebook`	Facebook	Use when the user provides a Facebook URL or asks to read personal posts, comments, reactions, friends, timelines, profiles, stories, feeds, groups, events, or saved items, or to discover public event
`facebook-cli`	Facebook	Use when the user provides a Facebook URL or asks to read personal posts, comments, reactions, friends, timelines, profiles, stories, feeds, groups, events, or saved items, or to discover public event
`figma`	figma	Work with Figma: read files, pages, frames and components, export assets, and manage design projects.
`flightaware`	FlightAware AeroAPI	Use for questions about a specific flight’s departure or arrival time, including “when’s my flight?” and confirmation of remembered times, plus flight status, delays, and cancellations. Verify the exa
`forget`	forget	Remove a personal fact, preference, relationship detail, topic, or prior event from Muse's active memory and stop existing copies or automations from bringing it back. Use for explicit requests such a
`function-health`	function-health	Retrieve lab biomarker results and clinician notes from Function Health.
`generate_podcast`	generate_podcast	Compose and deliver audio content: a podcast episode, briefing, or narrated summary, with one or more voices, as an MP3. For reading supplied text aloud verbatim, use tts.
`ghl`	ghl	Use HighLevel contacts, pipelines, appointments, messages, and its broader operation catalog.
`github`	GitHub	Search and work with the user's GitHub repositories through GitHub's official MCP server.
`gmail`	gmail	Work with the user's Gmail: search, read threads, draft, send, reply, forward, unsubscribe from mailing lists, manage labels, and open attachments.
`goals`	goals	Guidance for helping users create and accomplish goals. Read it before you create a goal for the user when no goal-creation contract is in context, and whenever you help with an existing goal. A Goals
`google-calendar`	google-calendar	Work with the user's Google Calendar: agenda views, event details, and scheduling changes.
`google-contacts`	google-contacts	Search, view, create, update, and delete the user's Google Contacts.
`google-docs`	google-docs	Read, create, and edit the user's Google Docs.
`google-drive`	google-drive	Work with the user's Google Drive: files, folders, uploads, downloads, and sharing.
`google-forms`	google-forms	Read, create, and update the user's Google Forms, and read responses.
`google-health-connect`	Health Connect	The user's synced Google Health Connect data from their Android device: daily metrics (steps, distance, calories, heart rate, HRV, VO2max), sleep sessions (stages, quality, efficiency), and workouts.
`google-sheets`	google-sheets	Read, write, and manage the user's Google Sheets.
`google-slides`	google-slides	Read, create, and edit the user's Google Slides presentations.
`google-tasks`	google-tasks	Manage the user's Google Tasks: lists, task details, creation, updates, and completion.
`granola`	granola	Search and read Granola meeting notes and transcripts through Granola's official public API (public-api.granola.ai). Requires a Granola API key (GRANOLA_API_KEY).
`healthex`	HealthEx	Use to connect HealthEx and ask questions about your medications, lab results, and other health records.
`idea-management`	Idea Management	Manage the Ideas tab: idea cards the agent can run, dismissing ideas, and explaining why an idea appeared or disappeared.
`image-search`	image-search	Search the web by text query for image URLs and source pages for feeds, artifacts, and visual references. Does not identify a supplied image or person.
`instagram`	instagram	Read Instagram profiles, followers, posts, comments, likes, stories, feed, saved content, and account insights. Answer questions about posts, reels, and Instagram links. Manage interests and profile d
`instagram-messages`	instagram-messages	Use this to interact with the user's Instagram messages. Read inboxes, threads, top recipients, filtered inbox views, DM search results, and send messages through `instagram-messages-cli`.
`klaviyo`	klaviyo	Run Klaviyo email/SMS marketing: manage lists and segments, build campaigns and flows, and edit templates.
`linear`	linear	Track engineering work in Linear: create and list issues, set priorities, manage cycles, and follow project status.
`lovable`	lovable	Lovable AI app builder. No public API exists, so this skill is catalog-only: the honest workaround is Lovable's GitHub sync, then operate on the code with the github skill.
`magic-moment`	magic-moment	Turn talking-head video footage into shareable short clips. The hosted AI-highlight feature has no public API, so this skill ships an honest local ffmpeg implementation: cut clips, burn in SRT captions, reframe to vertical 9:16.
`media-library`	media-library	Search and inspect the user's photo library, including connected device galleries. Use for photo requests and whenever a photo could ground or personalize a response; lookups of uploaded photos are ch
`messaging-channels`	Messaging Channels	Work across connected messaging providers (e.g. WhatsApp): check connection status, read side chats, send messages with approval.
`messenger`	Messenger	Work with the user's Messenger account: read call history; read and search contacts; read, search, and summarize conversations; send, react to, unsend, or edit messages; and message Marketplace listin
`meta-ads`	Meta Ads	Create, write, or manage Meta ads and assets: ad copy, campaigns, spend, reports, audiences, catalogs, product feeds, feed refresh schedules, experiments, and policy. Always load for any request to cr
`meta-threads`	meta-threads	Read and manage the user's Threads account: profile, posts, feed, saved posts, activity, insights, social graph, search, trends, and a specific post by URL or ID. Can tune feed ranking and publish pos
`muse-early-access`	Muse early access	Use for questions about Muse's general early access program, requests to join it, checking or withdrawing a join request, and admission updates.
`muse-feedback`	Muse feedback	Use for feedback and feature requests to the Muse team. Whenever your response tells the user you can't do a specific thing they wanted, or accepts them giving up on one, offer once to file feedback i
`muse_db`	muse_db	Inspect database-backed Muse records for diagnosis and cross-table tracing when purpose-built product tools do not expose the needed state.
`notion`	notion	Search, read, create, and update Notion pages via the Notion MCP.
`opentable`	OpenTable	Find restaurants on OpenTable, check availability, and make, change, or cancel reservations. Use for restaurant booking and live reservation data.
`outlook-calendar`	outlook-calendar	View, create, update, and delete events in the user's Outlook Calendar.
`outlook-contacts`	outlook-contacts	List, search, create, update, and delete contacts in the user's Outlook account.
`outlook-mail`	outlook-mail	Read, search, send, reply to, and delete messages in the user's Outlook Mail.
`paired-devices`	Paired Devices	Manage the user's paired devices: list devices, run commands, pull data such as location, and unpair devices.
`peloton`	peloton	Connect to Peloton to browse fitness classes, check schedules, and book workouts.
`permission-model`	Permission Model	Work with the runtime permission system: list pending requests, explain the access each grants, respect approve/deny decisions.
`personal-feed`	Personal Feed	Manage the user's personal Feed: editorial posts written on a schedule, the feed brief, regenerating or removing posts.
`philips-hue`	philips-hue	Control Philips Hue smart lights, rooms, scenes, and devices via the Hue Remote API v2.
`places-search`	Places Search	Find, compare, and share details on physical places near the user or in a specified area, including restaurants, cafes, bars, hotels, parks, attractions, shops, and businesses with local services. Not
`plaid`	Finances (Plaid)	Use to connect Plaid and read linked financial accounts: metadata, balances, transactions, recurring transactions, liabilities, and investments.
`podcast`	podcast	Compose and deliver audio content: a podcast episode, briefing, or narrated summary, with one or more voices, as an MP3. For reading supplied text aloud verbatim, use tts.
`printify`	printify	Use Printify to browse catalog data, manage shops and products, and review or create orders.
`quickbooks`	quickbooks	Read and manage the user's QuickBooks business through Intuit's official MCP server, including reports, invoices, customers, products, payment links, sales settings, and industry benchmarks.
`replit`	replit	Replit cloud IDE. Replit's public REST API is deprecated and its replacement is unreleased, so this skill is catalog-only: the honest workaround is pushing the Repl to GitHub, then operating on the code with the github skill.
`secure-vault`	Secure Vault	Handle credentials securely: collect passwords, API keys, and tokens only through secure entry UI, never in chat.
`self-awareness`	self-awareness	Ground self-referential answers in the agent's actual filesystem. Use when the user asks who the agent is, what it can do, what it knows, what it remembers, what it has built, what services are connec
`share-ideas`	Share ideas	Publish a portable Idea card from the Ideas tab — only after the user explicitly asks to publish it.
`shopify`	shopify	Manage a Shopify store: list and update products, orders, customers, inventory levels, and discount codes.
`shopping`	shopping	Use for any product or shopping question: find, reverse image search, shopping Instagram/Marketplace links, buy, compare, or evaluate real products with prices, images, and product page URLs, includin
`skill-creator`	skill-creator	Create or update a workspace skill: its description, structure, instructions, and supporting files.
`slack`	slack	Work with Slack: list channels, read and send messages, manage threads and reactions, and search history.
`social-content-performance`	Social Content Performance	Analyze the user's own Instagram account and post performance using linked-account analytics.
`spotify`	spotify	Discover, search, and manage Spotify music, podcasts, and playlists, including deleting shows or episodes you created with Save to Spotify.
`stripe`	stripe	Handle Stripe payments: create payment links and checkout sessions, manage customers, subscriptions, invoices, and refunds.
`subscription-status`	subscription-status	Answer questions about the user's Muse subscription, plan, usage, tokens, reset timing, or available plans and prices, or verify information that references the Muse subscription.
`tailscale`	Tailscale	Set up Muse's built-in Tailscale connector, join a tailnet or Headscale network, check status, and reach private machines through the TCP tunnel proxy. Read for Tailscale, VPN, MagicDNS, network egres
`tessie`	tessie	Monitor a Tesla vehicle, inspect live state, and run explicit Tessie command endpoints.
`threads`	threads	Read and manage the user's Threads account: profile, posts, feed, saved posts, activity, insights, social graph, search, trends, and a specific post by URL or ID. Can tune feed ranking and publish pos
`threads-messages`	threads-messages	Use this to interact with the user's Threads messages: read inboxes and message threads, and send messages through `threads-messages-cli`.
`ticketmaster`	ticketmaster	Search Ticketmaster events and seats with pricing. Returns Buy-now links to Ticketmaster checkout; it cannot complete a purchase itself.
`todoist`	todoist	Read and manage Todoist tasks, projects, comments, labels, filters, and reminders through Todoist's official MCP server.
`travel-planning`	travel-planning	Use this skill when an active or proposed trip needs planning, logistics, feasibility, entry or transit checks, itinerary work, or investigation of an airport process, immigration, ground transport, a
`tts`	tts	Turn supplied text into spoken audio, single or multi-speaker. For composed audio content (a podcast, briefing, or narrated summary), use podcast.
`vercel`	vercel	Deploy with Vercel: list projects and deployments, promote to production, manage domains and environment variables.
`voice-calls`	voice-calls	Provides the static system voice catalog used by Jarvis. It does not define a user-facing workflow.
`voice-design`	voice-design	Choose or design a speaking voice when the user asks for a new, different, custom, invented, or generated voice, or restore the voice used immediately before the current one.
`voice-selector`	voice-selector	Provides the static system voice catalog used by Jarvis. It does not define a user-facing workflow.
`wallet`	Wallet	Coordinate payments: check wallet connection state, use saved payment methods through secure provider pages, run purchase review and approval.
`wearable-device-skills`	Wearable Device Skills	Use when the user asks to discover, inspect, or invoke an agentic capability dynamically published by a paired phone or wearable, including device controls, app actions, camera or media actions, and s
`wearables-comms`	Wearables Calls and Messages	Handle calls and messages through connected wearables: read notifications, place and answer calls, and send quick replies.
`wide-research`	wide-research	Use when the user needs broad parallel research across many independent inputs with a shared output schema.
`withings`	withings	Use when linking Withings or reading Withings body measurements, activity, sleep, workout, heart, and intraday data.
`zapier`	zapier	Connect Muse to actions across apps through Zapier's official MCP server.
`zoom`	zoom	Manage Zoom: schedule and list meetings, fetch recordings, and manage meeting settings.

How to use

Browse the table above to see capability patterns.
Each Muse skill follows a pattern: clear purpose, structured input, and safe actions (read vs write separated).
If you want to build a Muse-like AI, copy the pattern: define skills as instructions + input schema, not just freeform prompts.

Contributing

PRs and issues are welcome. If you adapt these skills for other LLMs, share it here.

License

This project is licensed under the GNU Affero General Public License v3.0
(AGPL-3.0-only) — see LICENSE for the full text.

In short: you may use, study, modify, and share this work, but any modified
version you distribute (including over a network) must remain open-source
under the same license, with copyright and attribution notices intact.

Muse-Skill-Hub