temporary-license-expert
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or safety bypass attempts were detected within the skill instructions or metadata.
- [NO_CODE]: The skill consists entirely of markdown-based instructional text and YAML metadata. It does not include, download, or execute any scripts, binaries, or automation tools.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill is designed to ingest and analyze user-provided legal documents and license agreements (referenced in the 'Workflow for License Review/Drafting' section).
- Boundary markers: The instructions do not define specific delimiters for untrusted input, relying on the agent's internal context management.
- Capability inventory: The skill does not utilize any tools or commands with file-system access, network capabilities, or code execution privileges.
- Sanitization: There is no explicit sanitization of ingested text, but the lack of executable capabilities makes this surface low-risk.
Audit Metadata