ayon-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@ayon-mcplist all projects"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
ayon-mcp
A lean, standardized MCP server for AYON (Ynput) — drive an AYON production server from any AI agent through one clean interface.
Fourth in a set of single-purpose tracker MCPs, all the same shape so an agent (or a migration) can speak to any of them interchangeably:
shotgrid-mcp · ftrack-mcp · kitsu-mcp · ayon-mcp
What it gives you (20 tools)
One generic CRUD family over AYON's entities —
find/get/create/update/delete(entity types:folder,task,product,version,representation).Schema / discovery —
list_projects,get_project,list_folder_types,list_task_types,list_statuses(with the canonical mapping),list_tags,get_attributes,list_addons,whoami.Typed helpers —
new_folder(any folder type),new_task,new_product,new_version,set_status.project_summary— a normalized, cross-tracker snapshot (counts + per-shot tasks with canonical statuses) in the exact same shape the other three MCPs emit, so hub tools (verify / migrate / audit) work on AYON for free. AYON's polymorphic folders are mapped onto sequences/assets/shots byfolder_type.
Related MCP server: MCP Server Template
What's standardized here (vs a raw AYON client)
One CRUD family instead of dozens of typed endpoints —
find("folder", project, {...})etc.A two-level
dry_runon every write —"plan"(client-side echo, contacts nothing) and"preflight"(resolves references + validates statuses against live data, returns a before→after diff and anok/would_failverdict — writes nothing). OptionalMCP_PLAN_LOG=/path.jsonlrecords every plan.Canonical statuses — AYON's per-project statuses (Not ready / In progress / Pending review / Approved…) are mapped to the shared
todo/wip/done/review/approvedset, so cross-tracker logic is uniform.The normalized
project_summarycontract — identical to shotgrid/ftrack/kitsu-mcp.
Install
pip install ayon-python-api fastmcpConfigure (env only — no secrets in source)
export AYON_SERVER_URL="http://your-ayon:5000"
export AYON_API_KEY="<a service / API key>" # create one in AYON ▸ user ▸ API keysAdd to Claude Code / any MCP client:
claude mcp add ayon \
-e AYON_SERVER_URL=$AYON_SERVER_URL \
-e AYON_API_KEY=$AYON_API_KEY \
-- python /path/to/ayon-mcp/server.pyPart of a tracker-MCP set — migrate between platforms
Because all five MCPs emit the same project_summary and accept a uniform tool surface, an agent with two
loaded can copy a project across trackers (read source → write target) with no bespoke script — the
clone-as-hub thesis as five shippable MCPs. AYON's own ayon-ftrack addon does exactly this kind of sync
internally; this MCP exposes AYON to the same agent-driven workflow.
AYON specifics handled
Polymorphic folder hierarchy — folders are Episode/Sequence/Shot/Asset (any nesting);
new_foldertakes afolder_type, andproject_summaryflattens them onto the cross-tracker shape.Product → Version → Representation publish model — first-class in the CRUD family.
Anatomy / attributes —
get_attributes(entity_type)+get_projectexpose the schema-as-data.
Built on the official ayon-python-api (ayon_api) + fastmcp. MIT. Credits Ynput for AYON
(AGPL server, open source).
Docs
📊 COMPARISON.md — side-by-side of the five trackers (ShotGrid · ftrack · Kitsu ·
AYON · NIM): data model, status vocabularies, and the migration incompatibilities to know about.
🧪 TESTING.md — how these servers are validated (live round-trip tests + two-level
dry-run checks).
Built by John Huikku · alienrobot.com
This server cannot be deployed
Maintenance
Related MCP Connectors
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
MCP server for Clipkit — gives AI agents a video toolbox via the Clipkit schema.
MCP server for progressive tool usage at any scale (see https://klavis.ai)
MCP server that lets AI assistants use all OneSchema features exposed via the public API.
Related MCP Servers
- FlicenseNot gradedqualityFmaintenanceA production-ready MCP server built with FastAPI, providing an enhanced tool registry for creating, managing, and documenting AI tools for Large Language Models (LLMs).34-
- FlicenseNot gradedqualityDmaintenanceA production-ready Python scaffold for building Model Context Protocol (MCP) servers using FastMCP. It provides a structured framework for developers and AI agents to rapidly develop, test, and manage custom tools and workflows.1-
- AlicenseNot gradedqualityDmaintenanceA production-grade MCP server designed for multi-tenant, authenticated, and observable AI agent systems, enabling secure tool execution across heterogeneous data sources.64MIT
- AlicenseNot gradedqualityDmaintenanceA lean MCP server giving LLM agents full access to the ShotGrid / Autodesk Flow Production Tracking API through 15 curated tools, including generic CRUD, schema discovery, and safe writes with a dry_run flag.1MIT