Skip to content

rbutinar/ade-ops / .agents/skills/fabric-pipeline-poll

fabric-pipeline-poll

Verified

Poll a Fabric Pipeline Run to Terminal State

/fabric-pipeline-poll — Poll a Fabric Pipeline Run to Terminal State

Status: preview. F2 Fabric workspace lifecycle suite, 2026-05-28. Source pattern from distributions/demo-claude/.../local/pipeline_poll.py (DDF demo run monitor).

What it does

Polls the status of a Fabric Data Pipeline run until it reaches a terminal state (Succeeded, Failed, Cancelled, Deduped, Completed) or a timeout expires. Surfaces the final status and, on Failed, the failure reason and the failed activity if Fabric exposes it.

Companion to /fabric-pipeline-run. Use stand-alone when:

  • You triggered the run elsewhere (UI, scheduled trigger, another script) and want to monitor from the agent
  • A prior --wait invocation timed out and you want to resume polling

When to use

  • After a /fabric-pipeline-run without --wait to bring back the terminal status
  • Picking up an in-flight run that started elsewhere
  • Debugging: re-poll a known run_id to retrieve the failure detail

Prerequisites

  • FabricConnector credentials
  • Workspace + pipeline + run_id known

Usage

/fabric-pipeline-poll --pipeline-id <id> --run-id <id> --env dev
/fabric-pipeline-poll --pipeline-id <id> --run-id <id> --workspace-id <id>
/fabric-pipeline-poll --pipeline-id <id> --run-id <id> --env dev --timeout 1800   # max seconds, default 1800
/fabric-pipeline-poll --pipeline-id <id> --run-id <id> --env dev --interval 15    # poll every N seconds, default 15
/fabric-pipeline-poll --pipeline-id <id> --run-id <id> --env dev --once           # single status check, no loop

Terminal states

StateTreated as outcome
Succeededok
Completedok (synonym; some Fabric stack versions return this)
Failedfail (with failureReason surfaced)
Cancelledfail
Dedupedfail (run was deduped because another identical was in flight)

Non-terminal: NotStarted, InProgress, Queued. Loop continues.

Pipeline summary

  1. Resolve workspace id
  2. Load project + credentials, build connector
  3. Loop until terminal OR timeout:
    • GET /v1/workspaces/{ws}/items/{pipeline_id}/jobs/instances/{run_id}
    • Parse status field
    • If terminal: break
    • If --once: break after first check
    • Else: sleep --interval seconds
  4. If terminal Failed: also fetch + surface failureReason / error field and (if available) the activity-level error from the run details
  5. Surface: pipeline name + run_id + start time + duration + final status
  6. ops.log: PIPELINE-POLL | <env or -> | fabric: ws=<id> pid=<id> run=<id> terminal=<status> | <ok|fail> (or partial for --once when not yet terminal)

Output example

Polling pipeline acme_medallion run a1b2c3...
  [  0s] status=NotStarted
  [ 15s] status=InProgress
  [ 30s] status=InProgress
  [ 45s] status=Succeeded

FINAL: Succeeded
Duration: 45s

On failure:

  [ 60s] status=Failed
FINAL: Failed
failureReason: Activity 'run_silver' failed: AnalysisException: Table or view 'workspace.default.bronze_part' is not a valid identifier.

Preview tracking — known unknowns

  1. Polling cost: a 30-minute pipeline polled at 15s interval = 120 API calls. Per-tenant Fabric API quotas exist but are generous (~600/min on capacity-backed workspaces). Document this caveat.
  2. Activity-level error: the top-level failureReason is sometimes opaque ("Activity X failed"). The full per-activity error requires a second call to /jobs/instances/{run_id}/activityRuns. V1 fetches it automatically on Failed only; V2 should surface it inline.
  3. Run completion vs. dataset refresh lag: pipeline Succeeded does NOT guarantee downstream semantic-model refresh completed. Refresh is a separate async job. Caller must poll the dataset refresh status separately (different API).

Status — promotion to stable

  1. 3+ polls completed across Succeeded + Failed + Cancelled terminal states
  2. Activity-level error retrieval tested on a real failure
  3. Timeout behavior verified (does not hang indefinitely)
  4. ops.log entries reviewed

ARGUMENTS: $ARGUMENTS

writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 322kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall