fix-errors

What problem does it solve? Unhandled errors reported through error telemetry dashboards often get patched at the crash site with guards or try/catch blocks, which masks the real defect and lets invalid data keep flowing through the system. This Skill guides you to trace errors back to the producer of invalid data and fix the actual root cause. ## Core Features & Use Cases - Stack Trace Analysis: Read stack frames bottom-to-top to trace data flow and locate the producer of malformed data rather than the consumer that crashes. - Error Message Enrichment: When the producer is not visible in the stack (e.g., cross-process IPC senders), enrich thrown errors with diagnostic context like data type, truncated values, and operation names so the next telemetry cycle reveals the source. - Error Construction Analysis: Locate and read the code that constructs an error before fixing it, since construction logic reveals trigger conditions, error subtypes, and whether the error is actionable. - Use Case: Given a telemetry issue showing URI.revive crashing on invalid input, instead of adding a typeof guard in revive, enrich the IPC handler error message to capture the invalid value and identify which sender passes malformed URI data. ## Quick Start Use the fix-errors skill to investigate this telemetry issue with the attached stack trace and error message, and propose a root-cause fix.