meta-ai-valuation-adjustments
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from various external project and financial records which could serve as vectors for malicious instructions.
- Ingestion points: Inputs include client files, project files, research notes, interviews, operating records, and accounting records as specified in SKILL.md.
- Boundary markers: The instructions do not define delimiters or specific "ignore embedded instructions" warnings for the ingested data.
- Capability inventory: The skill contains no executable scripts and explicitly mandates a read-only posture, prohibiting network access and file modifications, which significantly limits the impact of potential injections.
- Sanitization: No explicit sanitization or validation logic is provided for the external inputs.
- [NO_CODE]: The skill consists entirely of markdown documentation and reference tables with no executable code, scripts, or automated tool invocations.
Audit Metadata