package-evaluator
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading the full content of package definition files (SKILL.md, AGENT.md, etc.) and reference files during its analysis phase.
- Ingestion points: Phase 2 of the workflow in SKILL.md explicitly instructs the agent to read definition files and reference files on disk for evaluation.
- Boundary markers: The instructions in SKILL.md do not specify the use of delimiters, XML tags, or 'ignore-embedded-instruction' markers when the content of these external files is processed.
- Capability inventory: No scripts are shipped with this skill; capabilities are limited to the file-reading and markdown reporting workflow described in SKILL.md. No network access, subprocess execution, or file-write operations are present.
- Sanitization: SKILL.md does not describe any sanitization or validation of the ingested text content to prevent instructions within the audited files from influencing the agent's audit behavior.
Audit Metadata