Documentation Index Fetch the complete documentation index at: https://docs.hellofriday.ai/llms.txt
Use this file to discover all available pages before exploring further.
In this guide, you’ll trigger a Friday workflow from your application, stream its progress in real time, and retrieve the results.
Prerequisites
Friday platform installed and running
At least one space loaded with a signal configured (Friday ships with starter spaces you can use)
The daemon API is available at http://localhost:18080
Friday ships with starter spaces for pull request code review and Jira bug fixes. You can use any of these to follow along — replace the workspace and signal IDs in the examples with your own. Run curl http://localhost:18080/api/workspaces to see what’s loaded.
Trigger a signal Every workflow in Friday starts with a signal . Your application sends a POST request with a payload, Friday runs the job, and returns results via JSON or a real-time SSE
Asynchronous request Send a POST request and get a JSON response when the job completes:
curl -X POST http://localhost:18080/api/workspaces/frozen_taco/signals/fix-bug \ -H 'Content-Type: application/json' \ -d '{ "payload": { "issue_key": "PROJ-123", "repo_url": "https://bitbucket.org/myteam/myrepo" } }'
Response: { "message" : "Signal completed" , "status" : "completed" , "workspaceId" : "frozen_taco" , "signalId" : "fix-bug" , "sessionId" : "sess_abc123" }
The request blocks until the job finishes. For long-running jobs, use SSE streaming or trigger the signal and poll for status separately.
Stream results in real time Add the Accept: text/event-stream header to get a real-time SSE stream as the job executes:
curl -N -X POST http://localhost:18080/api/workspaces/frozen_taco/signals/fix-bug \ -H 'Content-Type: application/json' \ -H 'Accept: text/event-stream' \ -d '{"payload": {"issue_key": "PROJ-123"}}'
SSE event stream: data: {"type":"data-session-start","data":{"sessionId":"sess_abc123"}} data: {"type":"text-delta","delta":"Reading Jira ticket..."} data: {"type":"job-complete","data":{"success":true,"sessionId":"sess_abc123","status":"completed"}} data: [DONE]
Intermediate events follow the Vercel AI SDK UIMessageChunk format (text-delta, tool-input-available, tool-output-available, etc.). Atlas adds custom data events prefixed with data- (e.g. data-session-start, data-agent-start). Check run status After triggering a signal, use the sessionId from the response to check on the run.
Get run details curl http://localhost:18080/api/sessions/sess_abc123
List runs for a space curl 'http://localhost:18080/api/sessions?workspaceId=frozen_taco'
Cancel a running job curl -X DELETE http://localhost:18080/api/sessions/sess_abc123
Retrieve artifacts Agents can produce artifacts — summaries, reports, database files, and other data generated during a run.
List artifacts curl 'http://localhost:18080/api/artifacts?workspaceId=frozen_taco&limit=10'
Get an artifact curl http://localhost:18080/api/artifacts/art_xyz
Export a database artifact as CSV Artifacts created from CSV uploads are stored as SQLite databases. Use the /export endpoint to download them back as CSV:
curl -o export.csv http://localhost:18080/api/artifacts/art_xyz/export
Upload a file curl -X POST http://localhost:18080/api/artifacts/upload \ -F 'file=@data.csv' \ -F 'workspaceId=frozen_taco'
Supported file types include CSV, PDF, DOCX, PPTX, images, audio, and plain text. CSV files are automatically converted to SQLite databases for use with the Data Analyst agent.
Send a chat prompt You can also send freeform prompts to a space. The response streams back via SSE.
curl -N -X POST http://localhost:18080/api/workspaces/frozen_taco/chat \ -H 'Content-Type: application/json' \ -d '{ "id": "chat-001", "message": { "role": "user", "content": "What signals are configured in this space?" } }'
Use the same id to continue a conversation across multiple messages.
Common patterns CI/CD: trigger a code review on every pull request Call the Friday API from a GitHub Actions workflow to review pull requests automatically:
.github/workflows/friday-review.yml
name : Friday PR Review on : pull_request : types : [ opened , synchronize ] jobs : review : runs-on : ubuntu-latest steps : - name : Trigger Friday review run : | curl -X POST "${{ secrets.FRIDAY_URL }}/api/workspaces/${{ secrets.FRIDAY_WORKSPACE }}/signals/review-pr" \ -H 'Content-Type: application/json' \ -d "{\"payload\": {\"pr_url\": \"${{ github.event.pull_request.html_url }}\"}}"
Set FRIDAY_URL to your daemon’s public URL (the webhook tunnel URL if running locally) and FRIDAY_WORKSPACE to your space ID.
Backend service: trigger and poll For backend services that need to trigger a job and wait for the result:
interface SignalResult { sessionId : string ; status : string ; } interface Session { status : "completed" | "running" | "failed" ; [ key : string ] : unknown ; } async function runFridayJob ( workspaceId : string , signalId : string , payload : Record < string , unknown > ) : Promise < Session > { // 1. Trigger the signal const triggerResponse = await fetch ( `http://localhost:18080/api/workspaces/ ${ workspaceId } /signals/ ${ signalId } ` , { method: "POST" , headers: { "Content-Type" : "application/json" }, body: JSON . stringify ({ payload }), } ); if ( ! triggerResponse . ok ) { throw new Error ( `Failed to trigger signal: ${ triggerResponse . status } ` ); } const { sessionId } = ( await triggerResponse . json ()) as SignalResult ; // 2. Poll for completion while ( true ) { const session : Session = await fetch ( `http://localhost:18080/api/sessions/ ${ sessionId } ` ). then (( r ) => r . json ()); if ( session . status === "completed" ) return session ; if ( session . status === "failed" ) throw new Error ( "Job failed" ); await new Promise (( r ) => setTimeout ( r , 2000 )); } } // Usage const result = await runFridayJob ( "frozen_taco" , "fix-bug" , { issue_key: "PROJ-123" , });
For production use, prefer SSE streaming over polling — it’s more efficient and gives you real-time progress.
Discover spaces and signals Before triggering signals, you may need to find which spaces and signals are available:
# List all spaces curl http://localhost:18080/api/workspaces # Get a space's full config (includes signals, jobs, agents) curl http://localhost:18080/api/workspaces/frozen_taco
The workspace response includes the signals object with each signal’s schema — use this to know what payload fields are expected.
Next steps
API reference Full endpoint reference with request and response schemas for every API operation.
Signals Learn about signal types — webhooks, cron schedules, Slack triggers, and more.
Spaces Understand how spaces organize agents, jobs, and signals into deployable units.
Jobs How jobs compose agents into multi-step pipelines with data contracts.
