What problem does it solve? Triaging a bug or change request normally means reading a long ticket thread, hunting through repositories, checking logs, and then writing up findings by hand. This Skill automates that end-to-end investigation and posts a verified analysis comment back to the work item, so engineers get an actionable root cause or approach instead of a raw ticket. ## Core Features & Use Cases - End-to-end triage workflow: Fetches the full issue thread and attachments, classifies the item as a bug or change request, investigates the relevant codebases with subagents, corroborates with observability data, and posts a structured analysis comment. - Tracker and observability adapters: Ships with Jira and GitHub Issues adapters plus Grafana and CloudWatch adapters, and degrades gracefully to any reachable platform via tool discovery. - Verification before posting: Re-checks the highest-confidence claims against actual files and line numbers before drafting, and supports a --dry-run mode that writes the comment to a file instead of posting. - Use Case: Given a Jira URL like acme.atlassian.net/browse/PROJ-123 reporting a slow export endpoint, the Skill reads the thread, mines the attached HAR file, traces the controller-to-database call path in the local checkout, confirms the missing pagination, and posts a root-cause comment with fix options. ## Quick Start Ask the AI to triage a work item by saying something like: triage https://github.com/myorg/myrepo/issues/42 and post what you find.
