bug-report-writer
Converts messy notes or failed test data into a formal, structured bug report for issue tracking systems.
Install
mkdir -p .claude/skills/bug-report-writer && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/17423" && unzip -o skill.zip -d .claude/skills/bug-report-writer && rm skill.zipInstalls to .claude/skills/bug-report-writer
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Use this skill whenever the user asks to write a bug report, turn notes into a Jira ticket, clean up a defect, make a bug report clear, summarize a failed test as a bug, draft a defect report, or create a QA issue from rough notes.Key capabilities
- →Produce a clear bug report from raw input
- →Format information into specific sections like Title, Summary, Environment
- →Suggest severity and priority with reasoning if not provided
- →Separate facts from assumptions and mark assumptions clearly
- →List open questions for missing or unclear information
How it works
This skill takes raw input and structures it into a bug report with specific sections such as Title, Summary, and Steps to reproduce. It also provides guidance on severity and priority.
Inputs & outputs
When to use bug-report-writer
- →Draft a Jira ticket from scratch using rough meeting notes
- →Format a failed test log into a formal defect report
- →Clarify ambiguous bug reports for developers
- →Standardize QA issues submitted by non-technical team members
About this skill
Take the user's raw input — rough notes, failed test observations, screenshot descriptions, reproduction steps, or a messy bug description — and produce a clear, professional bug report.
Output structure
Write the report using exactly these sections in this order:
Title One line. Specific and action-oriented. State what breaks, where, and under what condition. Avoid vague titles like "Login doesn't work."
Summary Two to four sentences. Describe the defect in plain language: what the user was doing, what went wrong, and what the impact is. Write for someone who has not seen the raw notes.
Environment Browser, OS, app version, device, or any other relevant context. If not provided, write: Not provided.
Preconditions The state the system must be in before the steps begin. Include account type, data setup, feature flags, or any other required setup.
Steps to reproduce Numbered. Each step is a single, testable action. Start from the earliest reliable point. If the steps are unclear or incomplete in the raw input, write what is known and flag gaps with: [unclear — needs verification].
Actual result What happened. State facts only. No interpretation. Quote error messages exactly if available.
Expected result What should have happened. Base this on the stated or implied requirement. If no requirement is stated, write a reasonable expectation and mark it: [assumed].
Severity Use one of: critical / major / minor / trivial.
- Critical — blocks core functionality or causes data loss; no workaround exists.
- Major — significantly impairs a feature but a workaround is possible.
- Minor — cosmetic or edge-case issue with low user impact.
- Trivial — typo, whitespace, or style inconsistency with no functional effect.
Do not use high, medium, or low. If severity is not provided, suggest one and explain the reasoning in one sentence.
Priority How urgently this needs to be fixed relative to other work. Use: P1 / P2 / P3 / P4. If not provided, suggest one and explain in one sentence. Priority and severity are not the same — a minor cosmetic issue on the home page seen by all users may be P2 even if it is severity minor.
Evidence Screenshots, logs, network responses, video links, or console output referenced in the raw input. List each item. If none were provided, write: Not provided.
Impact Who is affected, how often, and what they cannot do as a result. Be specific: "All users who attempt to reset their password via the email link" is better than "users."
Suggested owner or area The team, component, or code area most likely responsible. If not clear from the input, make a reasonable suggestion and mark it: [suggested].
Open questions List anything that is unclear, missing, or needs to be verified before the report can be considered complete. Write each as a question. If nothing is missing, write: None.
Rules
- Separate facts from assumptions. Mark assumptions clearly with [assumed] or [suggested].
- If information is missing, write "Not provided" — do not invent details.
- Keep reproduction steps numbered, minimal, and testable. Remove steps that are not needed to reproduce the issue.
- Use clear QA language. Active voice. No buzzwords or filler.
- Return plain readable text by default. If the user asks for a Jira-ready version, format the output in markdown with headers and code blocks where appropriate.
When not to use it
- →When the task is not related to writing or clarifying a bug report
Limitations
- →Requires raw input to generate a bug report
- →Does not invent details if information is missing
How it compares
This skill standardizes bug reports into a consistent format, ensuring all necessary details are captured and presented clearly, unlike an unstructured report.
Compared to similar skills
bug-report-writer side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| bug-report-writer (this skill) | 0 | 26d | No flags | Beginner |
| linear-ticket | 1 | 6mo | No flags | Intermediate |
| tpp | 0 | 5mo | No flags | Intermediate |
| acceptance | 0 | 1mo | No flags | Advanced |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
linear-ticket
useautumn
Refine rough engineering thoughts into structured Linear tickets with GitHub permalinks
tpp
photostructure
Work on a Technical Project Plan. Use when starting or continuing work on a TPP from _todo/.
acceptance
kirich1409
>
code-review-checklist
vudovn
Code review guidelines covering code quality, security, and best practices.
openspec-verify-change
studyzy
验证实现是否与变更产出物匹配。当用户想要在归档前验证实现是否完整、正确且一致时使用。
superpowers-finish
anthonylee991
Finalizes work: runs verification, summarizes changes, notes follow-ups, and ensures repo hygiene. Use at the end of an implementation or debugging session.