Cotana Hack Tracker
Detection Methodology
Evidence-first behavior scoring for incident classification and confidence evolution.
What Cotana Detects
- Wallet drainer behavior and abnormal outflow patterns.
- Protocol exploit signatures from contract balance deltas and method anomalies.
- Bridge exploit markers including unbacked mint and liquidity cliff behavior.
- LP exploits: exploits that drain value directly from a liquidity pool or AMM by manipulating pool mechanics, balances, or withdrawal paths.
What Cotana Does Not Detect
- No private off-chain intelligence or non-public incident feeds.
- No protocol-specific hardcoded exploit labels.
- No manual human narrative injection in generated summaries.
- No certainty guarantees; confidence depends on visible on-chain evidence quality.
Confidence Scoring
Confidence is behavior-derived from independent rule hits, interaction weighting between correlated vectors, and lifecycle updates over time. Scores rise when multiple high-signal vectors align in the same window and actor graph.
Summaries and timelines are generated deterministically from stored incident data and link directly to supporting on-chain evidence.