integration-management
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection Surface: This skill processes external, potentially untrusted documents such as purchase agreements and contract lists. This warrants review because instructions hidden in these documents could potentially influence the agent's actions or data processing. * Ingestion points: Accesses files from local paths (~/.claude/...), manual uploads, and connected repositories. * Boundary markers: No specific delimiters are defined to separate external document content from the skill's core instructions. * Capability inventory: Includes file read/write operations and CSV/table export capabilities. * Sanitization: Features a dedicated mechanism to neutralize potential formula injection in exported files.
- Sensitive Data Handling: The skill manages confidential M&A information. (1) It interacts with deal-specific metadata and tracking files. (2) This is a consideration as it involves the centralization of sensitive corporate data, requiring appropriate environment permissions.
Audit Metadata