OpenParable
Skip to content
← Back to skills

Sw Triage

Testing & Quality
sw-triage

Sw Triage is an Agent Skill for testing & quality published in the shopware/shopware repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Triage a Shopware 6 GitHub bug issue. Read the issue body, identify the affected code area via rg/git/gh, check for related fixes or duplicates, then emit a Markdown summary with disposition, severity, suggested domain labels, confidence, reasoning, and supporting evidence. Use when the user asks to triage, classify, label, or assess a Shopware issue, when they reference an issue by number (e.g. "#16599"), or when a new issue arrives that needs an initial dispositioning.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/shopware/shopware/blob/HEAD/.agents/skills/sw-triage/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/sw-triage, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Shopware Issue Triage

Context (interactive)

You operate inside the shopware/shopware monorepo with full read access to the codebase and to GitHub via shell tools. You cannot label, close, assign, or comment on the issue — your Markdown summary is the deliverable, the user decides what to do with it.

This skill drives the interactive triage path (Claude Code / opencode / Codex CLI in the repo). The unattended CI path runs in GitHub Agentic Workflows (gh aw) and uses a parallel policy fragment at .github/aw/sw-triage-policy.md that emits JSON via upload-artifact instead of Markdown. Both modes load the same shared policy from .github/aw/shared/sw-triage-policy.md (role, trust boundaries, research workflow, tool budget, anti-reward-hacking) so they cannot drift on the rubric.

Invocation

The user typed something like "triage issue #16599". Apply the shared policy in .github/aw/shared/sw-triage-policy.md — start with its Step 0 (fetch the issue) and continue through Step 6 (emit).

Output format

Emit a human-readable Markdown summary as your single final message. No JSON, no code fence. Structure:

  • H2 headline: ## Triage — Issue #<N>: <one-line defect description>
  • Decision table: Disposition, Severity, Confidence, Suggested labels, Duplicate of, Change size
  • Reasoning (2–5 sentences, opens with the one-sentence defect restatement from Step 1)
  • Evidence (bullet list, each prefixed [issue] or [shell])
  • Related work (Affected paths, Related PRs, Recent commits in area)
  • Missing template fields (or "none")

See assets/examples.md for the rendered JSON shape; the Markdown layout mirrors it field-for-field.

Final instruction

Apply the policy in .github/aw/shared/sw-triage-policy.md, then emit your Markdown summary using the structure above. The Markdown is your only output.