What problem does it solve? Triaging incoming bug reports manually is slow and inconsistent: engineers must read the report, hunt for the relevant code, judge severity, and draft a fix plan before any work begins. This Skill automates that triage by ingesting a bug report (pasted text or a URL), locating suspected code paths, assigning a verdict and severity, and writing a standardized assessment file that downstream fix and test commands consume. ## Core Features & Use Cases - Bug report ingestion: Accepts pasted text, URLs (GitHub, GitLab, Jira, Sentry, Stack Overflow), or a mix, with a strict URL trust policy that refuses private, loopback, and metadata endpoints. - Codebase triage: Searches the repository for symbols, error strings, and file paths mentioned in the report, then produces a root-cause hypothesis with a confidence level. - Structured output: Writes .specify/bugs/<slug>/assessment.md containing verdict, severity, reproduction steps, suspected code paths, proposed remediation, risks, and open questions. - Use Case: A maintainer pastes a GitHub issue URL reporting a login timeout; the Skill fetches the issue, locates the auth handler code, rates it high severity, and writes an assessment ready for the /speckit-bug-fix step. ## Quick Start Ask the assistant to assess a bug by providing the issue URL or pasting the error text, for example: assess this bug report https://github.com/org/repo/issues/42.
