tx-observability-traces

name: "tx-observability-traces" description: "Inspect queue health, diagnostics, dashboards, run traces, and recent failures. Use when working in Claude Code and the user needs tx commands from this area." metadata: short-description: "Inspect queue health, diagnostics, dashboards, run traces, and recent failures."

tx Observability And Traces

Use when the user is debugging tx behavior, stalled runs, or dashboard/API status.

Quick Start

• tx diag doctor
• tx trace list
• tx trace errors

Included Commands

• tx dashboard: Start API server + dashboard and open in browser
• tx diag: Diagnostics
• tx diag dashboard: Start API server + dashboard UI
• tx diag doctor: Run system health checks
• tx diag stats: Show queue metrics and health overview
• tx doctor: System health checks
• tx stats: Show queue metrics and health overview
• tx trace: Execution tracing for debugging run failures
• tx trace errors: Show recent errors across all runs
• tx trace heartbeat: Update run heartbeat state
• tx trace list: Show recent runs with event counts
• tx trace show: Show metrics events for a run
• tx trace stalled: List or reap stalled running runs
• tx validate is a deprecated alias for 'tx doctor'.

Full Help

Read references/commands.md for the full generated CLI help text for this skill's commands.

Search And Shell Fixes

When working in Claude Code, prefer rg -n <pattern> <path> and rg --files <path> over broad grep -r or fragile find pipelines.

If a shell/search command fails because of malformed flags, truncated paths, or broken quotes:

• rerun it as a smaller rg command with an explicit directory
• avoid partial paths like node_modul or unterminated quotes
• replace grep -r with rg -n unless rg is unavailable
• replace broad find probes with rg --files when you are really locating source files