soia-dev-run-ops-inspection
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns detected. The skill instructions prioritize security by enforcing read-only checks and forbidding the storage of sensitive data like tokens or cookies in reports.
- [EXTERNAL_DOWNLOADS]: The skill provides installation commands for packages hosted under the official vendor organization 'soia-team'. This is standard behavior for the vendor's ecosystem.
- [INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection as it ingests data from external sources such as metrics and status pages.
- Ingestion points: Data is collected from external monitoring sources and status pages defined in the checklist and configuration file.
- Boundary markers: The skill is instructed to distinguish provided facts from items requiring verification and to record raw observations without fabricating health baselines.
- Capability inventory: Capabilities include writing logs to a user-defined 'archive_root' directory and delegating diagnostic tasks to the 'soia-dev-terminal-ops' skill.
- Sanitization: The skill mandates the use of 'desensitized summaries' and 'desensitized evidence' to prevent the accidental exposure of sensitive information.
Audit Metadata