The CueAPI MCP server connects any MCP-compatible host to the CueAPI platform over stdio transport. It allows AI agents to schedule work and report write-once outcomes through CueAPI without being tied to a specific runtime environment. Configuration requires two environment variables: CUEAPI_API_KEY to authenticate with the CueAPI service, and CUEAPI_BASE_URL to point the server at the correct CueAPI endpoint. The package is distributed via npm under the identifier @cueapi/mcp at version 0.1.3, published by ai.cueapi, and the source is available on GitHub at cueapi/cueapi-mcp. Because the tool list exposed by this server is not documented in the current registry record, the precise operations an agent can invoke are not confirmed beyond the stated scheduling and write-once outcome reporting capabilities. This server is appropriate for workflows where an agent needs to queue tasks and record immutable results back to CueAPI. It is not suited for use cases requiring a different transport such as HTTP or SSE, and it depends on having valid CueAPI credentials and a reachable CueAPI endpoint before it will function.
Mcp
Schedule agent work and report write-once outcomes via CueAPI from any MCP host.
Use cases
- Schedule deferred agent tasks through a CueAPI-connected MCP host
- Report write-once outcomes from an agent run back to CueAPI
- Integrate CueAPI scheduling into any stdio-compatible MCP environment
- Automate task queuing pipelines that require immutable result recording
When to use it
- When an agent needs to schedule work via CueAPI from an MCP host
- When immutable, write-once outcome reporting to CueAPI is required
- When the runtime environment supports stdio transport for MCP
When not to use it
- When HTTP or SSE transport is required instead of stdio
- When no valid CUEAPI_API_KEY or reachable CUEAPI_BASE_URL is available
- When the use case requires reading or mutating previously recorded outcomes
- When detailed tool-level capability documentation is needed before adoption, as the tool list is not currently published