Dispatch flow

Dispatch turns a queued idea into a worker run. Treat it as a guarded operation, not a raw script call.

1. Check health and status

curl -fsS "$ENOCH_BASE_URL/healthz"
curl -fsS -H "Authorization: Bearer $ENOCH_CONTROL_TOKEN" \
  "$ENOCH_BASE_URL/control/api/status" | python3 -m json.tool

Review dispatch_safe and dispatch_blockers before proceeding. The operator dashboard UI is at http://<control-vm>:8787/control/dashboard-v2. The older /control/dashboard and /dashboard URLs redirect to the canonical V2 shell on current deployments.

The dashboard image in this section is a redacted, public-safe reference view with sample values. It shows the current V2 information architecture without exposing private hostnames, tokens, project IDs, or live operator data.

Current operator overview showing attention, running work, paper-writing, and publish/import cards

Use the dashboard as a fast visual check, but trust the API response as the source of truth for automation.

2. Check queue health

curl -fsS -H "Authorization: Bearer $ENOCH_CONTROL_TOKEN" \
  "$ENOCH_BASE_URL/control/api/queue-health" | python3 -m json.tool

3. Verify worker preflight

curl -fsS -H "Authorization: Bearer $ENOCH_CONTROL_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"wake_gate_url":"http://<worker>:8787","bearer_token":"<worker-token>","require_paused":false,"strict":false}' \
  "$ENOCH_BASE_URL/control/api/preflight" | python3 -m json.tool

4. Dry-run dispatch

curl -fsS -H "Authorization: Bearer $ENOCH_CONTROL_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"dry_run":true,"requested_by":"operator"}' \
  "$ENOCH_BASE_URL/control/dispatch-next" | python3 -m json.tool

Dry run should be your default for new config, new worker hosts, and post-upgrade checks.

5. Live dispatch

Only after status, queue health, and preflight are understood should you send a non-dry-run dispatch request. Confirm live_dispatch_enabled and dispatch_script_path are intentionally configured.

curl -fsS -H "Authorization: Bearer $ENOCH_CONTROL_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"dry_run":false,"requested_by":"operator"}' \
  "$ENOCH_BASE_URL/control/dispatch-next" | python3 -m json.tool

What the control plane checks

The queue tabs show the same read model from different operator angles. Empty active and queued views are useful during smoke testing because they prove the dashboard can read canonical queue state without launching work.

The two queue-tab images below are redacted, public-safe reference views with sample values. Use them to understand the queue layout, not as live queue evidence.

Current active queue tab showing bounded active-work rows

Current queued queue tab showing bounded ready-queue rows

The control plane checks that a queue item exists, pause and maintenance flags allow work, no conflicting active lane exists, worker preflight passes, and the dispatch script can be launched for live dispatch.

Get Started

Core Concepts

Guides

Configuration

Reference

Dispatch flow

Dispatch flow

1. Check health and status

2. Check queue health

3. Verify worker preflight

4. Dry-run dispatch

5. Live dispatch

What the control plane checks

​Dispatch flow

​1. Check health and status

​2. Check queue health

​3. Verify worker preflight

​4. Dry-run dispatch

​5. Live dispatch

​What the control plane checks

Dispatch flow

1. Check health and status

2. Check queue health

3. Verify worker preflight

4. Dry-run dispatch

5. Live dispatch

What the control plane checks