pipeline-investigate

Aggregate timelines, Docker logs, and decision events to diagnose pipeline run failures.

Updated Sep 22, 2025
One-click install
npx skills add https://github.com/tinkermonkey/switchyard --skill pipeline-investigate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pipeline-investigate
Source: https://github.com/tinkermonkey/switchyard/tree/main/.claude/skills/pipeline-investigate
Command: npx skills add https://github.com/tinkermonkey/switchyard --skill pipeline-investigate

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Investigate a pipeline run to quickly identify bottlenecks, failures, and root causes by collecting and correlating timeline data, Docker logs, and decision events.

Core Features & Use Cases

  • Timeline assembly: correlate events across stages to form a coherent run narrative.
  • Logs and decisions: fetch Docker logs and decision-event history to diagnose issues and verify progress.
  • Use Case: When a pipeline run stalls or fails, automatically gather metadata, timelines, and logs into a consolidated root-cause report for engineering teams.

Quick Start

Invoke the pipeline-investigate workflow with a valid pipeline_run_id to generate a timeline and root-cause report.

Frequently Asked Questions about pipeline-investigate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I investigate a stalled Docker pipeline run and find the root cause?

To investigate a stalled Docker pipeline run, you can aggregate timeline data, Docker logs, and decision events to correlate failures across stages and generate a consolidated root-cause report.

What is the best way to correlate decision events and logs for pipeline post-mortem analysis?

Correlating decision events and logs for pipeline post-mortem analysis involves fetching decision-event history and Docker logs, then assembling them into an end-to-end timeline to verify progress and diagnose issues.

Do I need an Elasticsearch cluster to diagnose orchestrated pipeline runs?

Yes, diagnosing orchestrated pipeline runs requires access to an Elasticsearch cluster with indices like pipeline-runs-*, decision-events-*, and agent-events-* to aggregate the necessary timeline and event data.

Can I use pipeline timeline assembly to debug failures across multiple orchestrated stages?

Yes, timeline assembly correlates events across multiple orchestrated stages to form a coherent run narrative, allowing you to quickly identify bottlenecks and debug failures in Claude Code pipelines.

What tools are used to fetch Docker logs and decision events for pipeline investigation?

Pipeline investigation uses curl, docker logs, and a Python timeline script to fetch metadata, logs, and decision events from a locally running orchestrator stack to produce the investigation results.

Why does my pipeline run fail to generate a root-cause report after a stall?

A pipeline root-cause report may fail if the locally running orchestrator stack is unavailable or if the Elasticsearch cluster lacks the required pipeline-runs-* and decision-events-* indices to aggregate timeline data.