ccf-integrity-auditor
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external artifacts including manuscripts, tables, and bibliographies. This creates a surface where adversarial instructions embedded within these documents could attempt to influence the agent's audit logic.\n
- Ingestion points:
SKILL.mdspecifies that the agent should identify and process supplied manuscripts, figures, tables, results, and bibliographies.\n - Boundary markers: There are no explicit markers or instructions defined to separate untrusted document content from the agent's operational logic.\n
- Capability inventory: The agent performs file reading, deterministic arithmetic cross-checks, and is authorized to update files in place for explicitly requested fixes.\n
- Sanitization: The skill does not implement sanitization or validation of the input text or metadata before processing.
Audit Metadata