AI Feedback Collector

Use this skill to turn free-form feedback about AI tool usage into a structured, objective issue report.

The goal is collection and normalization, not troubleshooting. Preserve the user's original meaning, avoid over-interpreting sparse descriptions, and make the output easy to paste into an issue tracker, spreadsheet, chat thread, or internal feedback system.

Workflow

1. Identify whether the user is reporting a problem with an AI tool or AI-assisted workflow.
2. Extract observable facts from the description: tool, task, scenario, failure behavior, impact, business context, and any available environment details.
3. Classify the likely problem category: model capability, environment/tooling, business-context clarity, workflow/process, user-skill/training, data/permission, safety/compliance, or unknown.
4. Separate facts from inferred possibilities. Mark uncertain fields as unknown instead of inventing details.
5. Normalize the report into the template below.
6. Add clear labels using the taxonomy in references/label-taxonomy.md when needed.
7. Include follow-up questions only as "Suggested additional information"; do not block output unless the user explicitly asks for an interview-style intake.

Output Template