| name | dalgo-pipelines |
| description | Run, monitor, create, and manage Dalgo orchestration pipelines (Prefect deployments), including the trigger-poll-verify loop. Use whenever a user asks to run a pipeline or check whether one succeeded. |
dalgo-pipelines
Use for pipeline workflows through Dalgo MCP tools.
Always
- Never stop at "triggered".
dalgo_trigger_pipeline_run only schedules a run — poll dalgo_get_flow_run to a terminal status before reporting the outcome.
- Resolve pipeline names to
deployment_id via dalgo_list_pipelines before any detail or mutation call.
- On
FAILED, CRASHED, or CANCELLED, fetch dalgo_get_flow_run_logs and quote the actual error before suggesting fixes.
- Confirm with the user before
dalgo_delete_pipeline; it is irreversible.
- If the user says "just trigger it, don't wait", trigger without polling but state explicitly that completion was not verified.
Decision Rules
- "Run pipeline X and make sure it worked": full trigger → poll → verify loop.
- "Did the last run succeed?":
dalgo_get_pipeline_run_history with limit=1 — do not trigger a new run.
- "Why did it fail?": route to
dalgo-troubleshooting.
- Terminal statuses:
COMPLETED, FAILED, CRASHED, CANCELLED. Non-terminal: SCHEDULED, PENDING, RUNNING.
- Poll every ~15 seconds, at most 20 times (~5 minutes); past that, report the run is still going and give the
flow_run_id for a later check.
Workflow Order
dalgo_list_pipelines → match by name → deployment_id.
dalgo_trigger_pipeline_run(deployment_id) → save flow_run_id; if the trigger errors, report and stop.
- Poll
dalgo_get_flow_run(flow_run_id) until terminal or poll limit.
COMPLETED: report success with timing. Failure: fetch logs, show the last ~30 lines, explain the error and a next step.
Retrieve