How do I install the Report Bug skill?

Run npx skillmds@latest add diegocortes15/report-bug in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Report Bug safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Report Bug?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Report Bug free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Report Bug?

Diegocortes15 (@diegocortes15) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Report Bug by diegocortes15 — Agent Skill

Name: Report Bug
Author: Diegocortes15

Report Bug

Turn a failed Playwright run into a ready-to-file bug report draft — repro steps, expected vs actual, correlated runtime observations and evidence paths — for a human to review and file. Never files anything itself.

Diegocortes15 a3a9dc5 5 files · 27.4 KB Updated Sep 17, 2026

File contents

report-bug

Given the last failed run, this skill assembles a bug report you can paste into a tracker: repro steps, the acceptance criterion the test traces to, expected versus actual, any runtime observations recorded during that test, and the paths to the screenshot, video and trace.

It files nothing. You invoke it, you read the draft, you decide. That division is the point: the tedious part is transcription, and the part that needs judgment — is this a defect in the application, or a ticket that describes behaviour the application never had? — stays with a person.

How to use it

Run the suite first, then:

Use the report-bug skill.

Or for one failure when several failed:

Use the report-bug skill for the test about sorting.

Workflow

The full procedural workflow is in references/workflow.md. Read that file before executing the skill.

References

references/workflow.md — the procedural workflow
references/report-template.md — the draft's structure, and the two-readings rule

Scripts

scripts/collect-failure.mjs — gathers the failure, its evidence, its acceptance criterion and its observations into structured JSON. Plain Node, no dependencies. Locating facts is lookup, not judgment, so it is a script rather than prose.
scripts/collect-evidence.mjs — copies that failure's screenshot, video and trace into one readable folder under bug-evidence/, with a README.txt carrying the error and how to open the trace. Copies only; the run output is never moved or deleted.

Scope

Filing into a tracker is deliberately not implemented. Jira writes are the exception, not the rule, in this project — only /refine-ticket writes, and only on explicit approval (ADR-0013). When a tracker write is added here it needs the same treatment and its own ADR.

ADR-0020 — why a run that cannot go green opens no PR; this skill is what you reach for afterwards
ADR-0021 — the runtime observations this report correlates
ADR index (origin repo) — rationale for every ADR-NNNN cited above. Paths like docs/… and src/… are relative to that repo, not to this skill.

---name: report-bugdescription: Turn a failed Playwright run into a ready-to-file bug report draft — repro steps, expected vs actual, correlated runtime observations and evidence paths — for a human to review and file. Never files anything itself.---# report-bugGiven the last failed run, this skill assembles a bug report you can paste into a tracker: repro steps, the acceptance criterion the test traces to, expected versus actual, any runtime observations recorded during that test, and the paths to the screenshot, video and trace.**It files nothing.** You invoke it, you read the draft, you decide. That division is the point: the tedious part is transcription, and the part that needs judgment — is this a defect in the application, or a ticket that describes behaviour the application never had? — stays with a person.## How to use itRun the suite first, then:> Use the report-bug skill.Or for one failure when several failed:> Use the report-bug skill for the test about sorting.## WorkflowThe full procedural workflow is in [`references/workflow.md`](references/workflow.md). Read that file before executing the skill.## References- [`references/workflow.md`](references/workflow.md) — the procedural workflow- [`references/report-template.md`](references/report-template.md) — the draft's structure, and the two-readings rule## Scripts- [`scripts/collect-failure.mjs`](scripts/collect-failure.mjs) — gathers the failure, its evidence, its acceptance criterion and its observations into structured JSON. Plain Node, no dependencies. Locating facts is lookup, not judgment, so it is a script rather than prose.- [`scripts/collect-evidence.mjs`](scripts/collect-evidence.mjs) — copies that failure's screenshot, video and trace into one readable folder under `bug-evidence/`, with a `README.txt` carrying the error and how to open the trace. Copies only; the run output is never moved or deleted.## ScopeFiling into a tracker is deliberately **not** implemented. Jira writes are the exception, not the rule, in this project — only `/refine-ticket` writes, and only on explicit approval (ADR-0013). When a tracker write is added here it needs the same treatment and its own ADR.## See also- ADR-0020 — why a run that cannot go green opens no PR; this skill is what you reach for afterwards- ADR-0021 — the runtime observations this report correlates- [ADR index (origin repo)](https://github.com/Diegocortes15/playwright-ai-framework/tree/main/docs/adr) — rationale for every `ADR-NNNN` cited above. Paths like `docs/…` and `src/…` are relative to that repo, not to this skill.

Frequently asked questions