eagle3-review-logs
Fail
Audited by Snyk on Jul 6, 2026
Risk Level: HIGH
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 0.90). The prompt explicitly instructs reading and echoing the tails of experiment log files and including quoted error lines in the report, which would cause the model to output any secrets (API keys, tokens, cookies, passwords) present in those logs verbatim.
Issues (1)
W007
HIGHInsecure credential handling detected in skill instructions.
Audit Metadata