Cueapi MCP server
{
"mcpServers": {
"cueapi": {
"command": "npx",
"args": [
"-y",
"@cueapi/mcp@latest"
],
"env": {
"CUEAPI_BASE_URL": "https://...",
"CUEAPI_API_KEY": "<cueapi_api_key>"
}
}
}
}{
"mcpServers": {
"cueapi": {
"command": "npx",
"args": [
"-y",
"@cueapi/mcp@latest"
],
"env": {
"CUEAPI_BASE_URL": "https://...",
"CUEAPI_API_KEY": "<cueapi_api_key>"
}
}
}
}{
"mcpServers": {
"cueapi": {
"command": "npx",
"args": [
"-y",
"@cueapi/mcp@latest"
],
"env": {
"CUEAPI_BASE_URL": "https://...",
"CUEAPI_API_KEY": "<cueapi_api_key>"
}
}
}
}{
"servers": {
"cueapi": {
"command": "npx",
"args": [
"-y",
"@cueapi/mcp@latest"
],
"env": {
"CUEAPI_BASE_URL": "${input:cueapi_base_url}",
"CUEAPI_API_KEY": "${input:cueapi_api_key}"
}
}
},
"inputs": [
{
"id": "cueapi_base_url",
"type": "promptString",
"description": "Cueapi base url",
"password": true
},
{
"id": "cueapi_api_key",
"type": "promptString",
"description": "Cueapi api key",
"password": true
}
]
}{
"mcpServers": {
"cueapi": {
"command": "npx",
"args": [
"-y",
"@cueapi/mcp@latest"
],
"env": {
"CUEAPI_BASE_URL": "https://...",
"CUEAPI_API_KEY": "<cueapi_api_key>"
}
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsoninstall npx -y @cueapi/mcpversion 0.1.3
What it does
An MCP server that connects Claude Desktop, Cursor, Zed, or other MCP hosts to CueAPI, a coordination layer for AI agent systems. It lets you schedule agent work, track execution history, and report outcomes with evidence from within a conversation.
- Schedule a cron job or one-time task to run agent work at a future time.
- Check execution history and fetch details on how past scheduled tasks completed.
- Claim and process pending work executions and report back outcomes with external IDs or URLs as proof.
Tools this server exposes
| Tool | Does |
|---|---|
| cueapi_create_cue | Create a recurring (cron) or one-time task. |
| cueapi_list_cues | List all cues, optionally filtered by status. |
| cueapi_get_cue | Fetch details for a single cue. |
| cueapi_fire_cue | Trigger an existing cue immediately with optional payload override. |
| cueapi_update_cue | Update a cue's name, schedule, callback URL, timezone, payload, or description. |
| cueapi_pause_cue | Pause a cue to stop it from firing. |
| cueapi_resume_cue | Resume a paused cue. |
| cueapi_delete_cue | Delete a cue permanently. |
| cueapi_list_executions | List historical task executions, optionally filtered by cue or status. |
| cueapi_get_execution | Fetch a single execution with its state and outcome. |
| cueapi_list_claimable_executions | List unclaimed worker executions, optionally filtered by task or agent. |
| cueapi_claim_execution | Claim a specific execution for processing. |
| cueapi_claim_next_execution | Claim the next available execution, optionally filtered by task name. |
| cueapi_execution_heartbeat | Extend the claim lease on an in-flight execution. |
| cueapi_report_outcome | Report a write-once outcome with evidence such as external ID or URL. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.