DEDUPED Hot AI repos, turned into use cases Get the Monday 5

How to publish a finished workflow as an authenticated HTTP API with Loom Flow

Expose a tested workflow as a POST endpoint that your app or script can call with one API key.

Short answer: Build and test the workflow on the canvas until one run succeeds; then Publish the workflow from the app, which auto-generates your global API key on first publish; copy the API key immediately, because it is shown only once. All steps.

Credentials. One global API key calls every published workflow, so store it in a secret manager and never commit it to code.

Who it is forA developer who wants to add an AI feature to an existing product and needs a stable endpoint instead of running the canvas by hand.

The problemA workflow that only runs inside a canvas cannot be called from other systems.

The outcomeA published workflow endpoint that returns results when called with a Bearer key from curl or another client.

What you need first

  • A machine with Docker and Docker Compose installed, plus git and a terminal (the README says 1 GB RAM is enough)
  • An API key for an LLM provider such as DeepSeek, or any OpenAI-compatible endpoint, added later through the admin UI
  • A web browser to open http://localhost:5000

Steps

  1. Build and test the workflow on the canvas until one run succeeds
  2. Publish the workflow from the app, which auto-generates your global API key on first publish
  3. Copy the API key immediately, because it is shown only once
  4. Call the endpoint with curl -X POST https://your-host/api/publish/{workflowId}/execute -H "Authorization: Bearer ffk_xxxxx" -H "Content-Type: application/json" -d '{"inputs": {"query": "..."}}' (inferred, check the README)
  5. If the key expires, regenerate it on the API Keys page

Prompt for your agent

Using the LoomFlow repo at https://github.com/banmu123/LoomFlow, publish the workflow <workflowId> and call its execute endpoint with the input <sample input>. Expected output: a JSON response from POST /api/publish/<workflowId>/execute. Verify that the call returns HTTP 200 with the workflow result and that the Authorization header uses your current API key.

Caveat. The README gives the endpoint pattern but not a full list of request fields, so check the inputs your workflow expects before you send the request.

Use this in your AI agent

Ask your agent

Using the Deduped MCP, get the setup steps for banmu123/LoomFlow and "Publish a finished workflow as an authenticated HTTP API"

Add Deduped to Claude Code once

claude mcp add --transport http deduped https://dedupedai.com/mcp

Cursor, ChatGPT, others

Share: X LinkedIn Hacker News Email Copy link

Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer