What problem does it solve?
Game-security research often mixes raw observations with unverified attribution, leading to overstated anti-cheat conclusions, uncalibrated detectors, and citations that do not actually support the claims made. This Skill enforces evidence-grounded reasoning so conclusions never exceed what the data supports.
Core Features & Use Cases
- Layered Reasoning Discipline: Separates observation, finding, attribution, and decision so an anomaly or invariant violation is never automatically treated as proof of cheating.
- Citation Verification & Claim Ledger: Verifies every URL, DOI, and source against authoritative metadata, then records each claim with its evidence, assumptions, counterevidence, and uncertainty label.
- Detector & Telemetry Evaluation: Guides calibration of anti-cheat detectors with proper data splitting, confusion matrices, prevalence reporting, and revalidation after patches or population drift.
- Use Case: When comparing two anti-cheat techniques described across README files and vendor posts, use this Skill to verify each source, label claims as observed or inferred, and conclude narrowly (supported, suspicious, or inconclusive) instead of overstating the evidence.
Quick Start
Ask the AI to evaluate whether the evidence in these game-security sources supports the claim that a specific anti-cheat technique detects DMA cheats, and to produce a claim ledger with verified citations.
