oma-backend
Pass
Audited by Gen Agent Trust Hub on May 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill incorporates stack manifests that define shell commands for compilation and testing (e.g., 'cargo check', 'bun test', 'uv run pytest') to be executed during the verification phase.\n- [CREDENTIALS_UNSAFE]: Rust code templates include a default fallback secret ('secret') for JWT decoding if no environment variable is provided. This is a common template pattern and is mitigated by the skill's explicit instructions to load production secrets from environment variables.\n- [DATA_EXFILTRATION]: The skill provides an error recovery playbook that suggests writing session progress to '/tmp' files in case of internal memory tool failures.\n- [EXTERNAL_DOWNLOADS]: The skill references and encourages the use of established, reputable backend frameworks and libraries such as FastAPI, NestJS, Axum, and Prisma.
Audit Metadata