What problem does it solve?
This Skill eliminates the risk of acting on unproven, vibes-based claims about the Cambridge Beer Festival app codebase, which commonly lead to misclassified data, regressions, and wasted time on incorrect review feedback.
Core Features & Use Cases
- Empirical data census: Count live API field types to validate parsing logic and identify dead code branches.
- Minified crash forensics: Decode Flutter web release crash stack traces to build exact, frame-by-frame root cause explanations.
- Async/race audit: Identify unawaited futures and shared state mutation risks that cause flaky, intermittent behavior.
- Review comment refutation: Construct tooling-backed proofs to dismiss incorrect automated or human review feedback.
- Staleness/retry semantics analysis: Model cache and network state machines to avoid retry storms or stale data lockouts for offline users.
- Hypothesis-prediction experiments: Force pre-experiment predictions to catch mismatches between assumed and actual system behavior.
A common use case is validating API field type claims before modifying parsing code, or decoding a minified production crash to implement a targeted, minimal fix.
Quick Start
Use the proof-and-analysis-toolkit skill to confirm whether the abv field in the live festival API is always a string before updating the drink parsing logic.
