oss-monetization
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting untrusted data from external software repositories, including license history, contributor lists, and issue tracker comments.
- Ingestion points: The agent is instructed to read project repository files, contribution records, and organizational mentions in issue trackers.
- Boundary markers: The skill includes clear directives to distinguish advisory content from legal or tax advice and mandates the use of specific reference documents.
- Capability inventory: The skill does not possess capabilities for remote code execution, network exfiltration, or dangerous system modifications; it is limited to providing analysis and storing task state in a local context file.
- Sanitization: No automated sanitization of ingested repository data is described.
- [SAFE]: All external skill references (e.g., samber/developer-relations-skills) belong to the same author as the analyzed skill, representing a trusted and consistent vendor scope.
- [SAFE]: Reference links in the documentation point to established technology news platforms and official project sites (Sidekiq, Laravel, GitHub, etc.) for educational and illustrative purposes only.
Audit Metadata