stellary-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., "@stellary-mcpshow my active 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.
Stellary Project Management MCP Server
Connect AI assistants and coding agents to live Stellary projects through the Model Context Protocol (MCP). The hosted server exposes project, board, collaboration, cockpit, and agent-mission tools while preserving Stellary's existing permissions.
Endpoint:
https://api.stellary.co/mcpTransport: Streamable HTTP
Authentication: OAuth 2.1 with PKCE and refresh-token rotation; bearer PATs remain supported for compatibility
Official MCP Registry listing · Documentation · Guide en français
What you can do
Discover projects, columns, cards, documents, and cockpit state.
Create and update cards, move work, assign teammates, and add comments.
Inspect sprints, priorities, agent status, and pending proposals.
Run governed agent missions with workspace rules and approval policies.
Use installed workspace integrations from eligible agent sessions.
An OAuth connection can authorize Me, one or more active workspace agents, or both. The exact tool list is resolved at connection time and checked again for the identity selected on every call. It depends on OAuth scopes, project access, workspace configuration, and each agent's tool and autonomy policies.
Related MCP server: MCP Server Toolkit
Connect in three steps
Add the hosted endpoint to an OAuth-capable MCP client.
Sign in to Stellary in the browser window opened by the client.
Choose a workspace and authorize Me, one or more active agents, or both.
Claude Code
claude mcp add stellary \
--transport streamable-http \
https://api.stellary.co/mcpCursor and JSON-based clients
{
"mcpServers": {
"stellary": {
"url": "https://api.stellary.co/mcp"
}
}
}A reusable example is available in
examples/mcp-client.json. Marketplace install
and PAT fallback scopes are documented below.
Install from the Cursor Marketplace
This repository is the Cursor Plugin wrapper for the hosted Streamable HTTP
endpoint. It does not start a local stdio or npx server.
Open Customize in Cursor and install Stellary from the Cursor Marketplace once the listing is live (submission: publish form).
Complete the Stellary OAuth window. Choose a workspace and authorize Me, one or more active agents, or both.
Ask Cursor to list your Stellary projects.
Until the listing is public, symlink this repo into
~/.cursor/plugins/local/stellary or add the same hosted URL to Cursor MCP
settings / ~/.cursor/mcp.json:
{
"mcpServers": {
"stellary": {
"url": "https://api.stellary.co/mcp"
}
}
}Cursor discovers OAuth from Stellary metadata. The Marketplace package therefore
ships no Bearer header and no ${STELLARY_TOKEN} placeholder.
Compatibility PAT (only if OAuth cannot complete)
Create a dedicated token from Account settings → API tokens. Start with
projects:read and pilotage:read. Add write scopes only when the client
needs them. Keep the secret in user MCP settings — never in this repository:
{
"mcpServers": {
"stellary": {
"url": "https://api.stellary.co/mcp",
"headers": {
"Authorization": "Bearer ${env:STELLARY_TOKEN}"
}
}
}
}The publish checklist for marketplace-publishing@cursor.com / the publish
form is in CURSOR_SUBMISSION.md.
Recommended first request
Ask the client to list your Stellary projects. This confirms authentication and project visibility without changing data. Then ask it to inspect one project's columns and cards before enabling write scopes.
Security
Revoke an OAuth connection from Workspace settings → MCP connections when a client should no longer have access.
Treat compatibility PATs like passwords and never commit one to a repository.
Start with read-only scopes and use an expiry date when a PAT is required.
All requests are still subject to Stellary permissions and rate limits.
Stellary never sends the client the private tokens attached to your agents.
For a vulnerability, follow SECURITY.md. For setup questions, email support@stellary.co.
Agent directory manifests
This repository is also the public discovery source for agent plugin
directories. OAuth-capable surfaces point at the hosted Streamable HTTP endpoint
without embedding a secret. Legacy PAT configuration remains documented for
clients that cannot complete remote OAuth. None starts a local stdio/npx
server.
Surface | Files |
cursor.directory / Open Plugins | |
Cursor Marketplace |
|
Claude Code Plugin Directory |
|
Gemini CLI Extensions | |
skills.sh / ClawHub |
|
Grok Build | |
GitHub Copilot / Agent Plugins | |
OpenAI Plugins / Codex |
|
Only set STELLARY_TOKEN when a client cannot use OAuth. Never commit a real
token. A 400×400 listing icon is in
assets/logo-400.png.
About this repository
This public repository is the canonical discovery and configuration source for Stellary's hosted MCP server. It contains the official MCP Registry manifest, client examples, documentation, and automated availability checks. The hosted Stellary application and server implementation are not distributed from this repository.
Thin plugin manifests for Grok Build and GitHub Copilot live in
.mcp.json, .grok-plugin/plugin.json,
and plugin.json. They point at the hosted endpoint
https://api.stellary.co/mcp (Streamable HTTP with OAuth, plus PAT
compatibility). A 400×400 listing icon is in
assets/logo-400.png.
The metadata and documentation in this repository are licensed under the MIT License. Use of the hosted service is governed by the Stellary terms.
This server cannot be deployed
Maintenance
Related MCP Connectors
AI agents use CalmBoard MCP to access board data, tasks, insights, and project workflows.
- Connext MCPOAuthcom.rti
Integrate Connext AI into agentic workflows, enabling Connext product-aware assistance.
- OneLoreOAuthai.onelore
Shared project context for AI agents and teams: docs, tasks, and messages that stay current.
Connect AI agents to ProductNow's context engine to search, create, review, and act.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceA comprehensive Model Context Protocol toolkit that transforms AI assistants into autonomous agents capable of executing real-world tasks across filesystems, web requests, Git workflows, databases, system commands, and AI integrations.MIT
- FlicenseNot gradedqualityDmaintenanceEquips AI coding agents with filesystem, Git, database, and computation tools via the Model Context Protocol.1-
- AlicenseNot gradedqualityAmaintenanceEnables secure enterprise AI agents to access internal tools like GitHub, Gmail, Calendar, file systems, databases, and knowledge bases through the Model Context Protocol, with built-in security, audit, and observability.MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to securely interact with local files, live internet search, databases, and development tools through the Model Context Protocol, turning them into autonomous production-ready assistants.5MIT