What problem does it solve? When production breaks—errors, outages, degraded performance—teams often guess at fixes and make things worse. This Skill enforces a disciplined SRE-style investigation workflow that finds the root cause with cited evidence before any fix is proposed. ## Core Features & Use Cases - Platform Detection & Evidence Gathering: Automatically detects Railway, Vercel, Docker, or Modal deployments, bootstraps missing CLIs, and launches parallel subagents to collect logs, deployment history, database state, and resource metrics. - Root Cause Analysis with Confidence Scoring: Presents findings in a structured table with High/Medium/Low confidence levels, each backed by specific log lines, metrics, or code references. - Scored Fix Plans & Documentation: Proposes fixes scored by Impact, Effort, and Risk, then generates investigation documents and numbered fix plans under documentation/planning/investigations/. - Use Case: Your API starts returning 500s after a deploy. Invoke the skill to pull Vercel error logs, trace the failing code path, identify the root cause with evidence, and receive a prioritized fix plan—without writing any code prematurely. ## Quick Start Ask the assistant to investigate why your production app is throwing errors or running slowly, and it will begin the evidence-driven debugging workflow.
