10-financial-projections
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, including client commercial records, contracts, and usage/cost evidence (file: SKILL.md).
- Ingestion points: External content is ingested via the 'Required Inputs' section and 'Generation Process' in SKILL.md.
- Boundary markers: Standard markdown headers and YAML frontmatter are present, providing some separation, but no specific 'ignore embedded instructions' warnings are applied to the external data blocks.
- Capability inventory: The skill refers to the execution of a local tool
formula_map.pyto audit Excel workbooks (file: SKILL.md). - Sanitization: The documentation does not specify any sanitization or validation routines for the external input content.
- [SAFE]: The skill references the execution of a local script
formula_map.pywithin the repository structure to perform automated auditing of financial workbooks. This is a legitimate development workflow and is documented as a quality control step (file: SKILL.md).
Audit Metadata