oe-workflow-triage

shami-ah's avatarfrom shami-ah

Workflow bug triage for OpenEvent. Use when you have a regression/bug report (wrong step, wrong detour, HIL stuck, site visit issues) and need a minimal reproduction trace plus a fix plan tied to workflow invariants and tests.

⭐0stars🔀1forks📁View on GitHub🕐Updated Jan 8, 2026
[Debugging]

When & Why to Use This Skill

The oe-workflow-triage skill is a specialized diagnostic tool designed for systematic bug triage and regression analysis within the OpenEvent ecosystem. It automates the transition from bug reports to verified fixes by generating minimal reproduction traces, validating workflow invariants, and structuring a phased PR ladder. This skill is essential for developers needing to resolve complex workflow issues like 'Human-In-The-Loop' (HIL) stalls and site visit regressions with high precision and minimal risk.

Use Cases

  • Reproducing complex workflow regressions by mapping bug reports to deterministic scenarios and generating execution traces.
  • Validating system state against workflow invariants to identify exactly where a step or detour failed during a site visit.
  • Creating new automated test scenarios for edge cases that are not covered by existing manual UX scripts to prevent future regressions.
  • Formulating a structured 1–3 PR fix plan that separates characterization tests from the actual code fix and subsequent cleanup for safer deployments.
nameoe-workflow-triage
descriptionWorkflow bug triage for OpenEvent. Use when you have a regression/bug report (wrong step, wrong detour, HIL stuck, site visit issues) and need a minimal reproduction trace plus a fix plan tied to workflow invariants and tests.

oe-workflow-triage

Minimal reproduction first (avoid “guess fixes”)

  1. Pick the closest deterministic scenario:

    • Site visit: scripts/manual_ux/manual_ux_scenario_I.py, scripts/manual_ux/manual_ux_scenario_H.py
    • Late changes / detours: scripts/manual_ux/manual_ux_scenario_E.py etc.

  2. Produce a trace:

    • python3 scripts/manual_ux/manual_ux_scenario_I.py > /tmp/repro.json

  3. Validate with invariants:

    • python3 scripts/manual_ux/validate_manual_ux_run.py /tmp/repro.json --require_site_visit

  4. If the bug is not covered by existing scenarios:

    • Create a new scripts/manual_ux_scenario_*.py by copying the closest scenario.
    • Keep it deterministic (AGENT_MODE=stub + explicit mapping + intent overrides).
    • Add/extend validator checks only when they prevent regressions.

Fix plan outputs

  • A 1–3 PR ladder:
    • PR1: characterization test / scenario reproduction
    • PR2: fix with minimal blast radius
    • PR3: cleanup/refactor (optional)