Model Context Protocol (MCP) Streamable HTTP + SSE Zero DevOps

Connect AI Assistants via MCP

Connect Claude.ai, Cursor, Antigravity IDE, Claude Code, Claude Desktop, and ChatGPT directly to LiteLambda. Give your AI coding agents the ability to write, deploy, inspect, and trigger production Python cron jobs autonomously.

Live Hosted MCP Server ONLINE MCP v2024-11-05
Streamable HTTP: https://mcp.litelambda.in/mcp
SSE Endpoint: https://mcp.litelambda.in/sse
Get Your API Token

Dynamic Configuration Generator

Paste your API token below to automatically customize every config snippet on this page:

Create Token ↗

1. Select Your Tool / IDE

Choose your AI platform to see tested, step-by-step setup instructions.

Cursor IDE Setup Recommended

Configure LiteLambda in Cursor via graphical settings or project-level config file.

Method A: Graphical Settings (Fastest)

1
Open Features Settings

Press Cmd + , (or Ctrl + ,), navigate to Features in the sidebar, and scroll down to MCP Servers.

2
Click "+ Add New MCP Server"

Name: litelambda
Type: select sse
URL: https://mcp.litelambda.in/sse

3
Add Authorization Header

Key: Authorization
Value: Bearer ll_live_YOUR_TOKEN

Method B: Project Config File (.cursor/mcp.json)

Create or edit .cursor/mcp.json in your project root to share MCP configuration with your team:

{
  "mcpServers": {
    "litelambda": {
      "url": "https://mcp.litelambda.in/mcp",
      "headers": {
        "Authorization": "Bearer ll_live_YOUR_TOKEN"
      }
    }
  }
}
Test it: In Cursor Chat (Cmd + L), type: "List my LiteLambda cron jobs".

2. Quick Connection Test (via Terminal)

Verify your API token and endpoint connectivity in 2 seconds using curl:

curl -s -X POST https://mcp.litelambda.in/mcp \
  -H "Authorization: Bearer ll_live_YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 1, "method": "tools/list"}'

3. Available Native AI Tools

Your connected AI assistant automatically receives these 7 native execution tools:

Tool Name What It Does Example Prompt for Your AI
list_jobs Returns all scheduled crons, active states, execution intervals, and last execution status. "What Python crons do I currently have running?"
deploy_job Deploys or updates Python script code, cron schedule expression, and timeout limit. "Deploy an hourly scraper to check competitor prices and post to Slack."
get_job Fetches script source code, pip package list, environment variables, and job metadata. "Show me the source code for my daily-report job."
trigger_job Triggers an immediate isolated run in a sandboxed container without waiting for the next cron interval. "Trigger the price scraper right now to test it."
get_logs Returns runtime stdout, stderr, execution duration, and exit code of recent runs. "Did the last run of the slack-bot succeed? Show me the logs."
pause_job Pauses or resumes regular scheduled executions with zero downtime. "Pause the hourly job while we update the database."
delete_job Permanently deletes a cron job and its historical execution logs. "Delete the old test scraper job."

4. Troubleshooting & FAQ

Q: Why are tools not appearing in my chat interface?

Most IDEs cache MCP tools upon startup. In Cursor, go to Settings > Features > MCP and click the reload icon 🔄. In Claude Desktop, restart the app completely with Cmd + Q. In Claude.ai, make sure the plug icon in the chat bar is toggled ON.

Q: Getting 401 Unauthorized errors?

Ensure your authorization token format is exactly Bearer ll_live_YOUR_TOKEN. Double check that there are no accidental trailing spaces or quotation marks in the token string. You can verify your token instantly using the curl test above.

Q: When should I use /mcp vs /sse?

Use https://mcp.litelambda.in/mcp for modern clients supporting Streamable HTTP (Claude.ai custom connectors, Claude Code CLI, modern Cursor). Use https://mcp.litelambda.in/sse for clients requiring Server-Sent Events (Antigravity IDE, Windsurf, mcp-remote).