skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/schedule-cost-link/Gen Agent Trust Hub
schedule-cost-link
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill integrates external data from CSV, Excel, and JSON files, which creates a potential surface for indirect prompt injection. Ingestion points: The agent is instructed in instructions.md to accept common file formats and user parameters. Boundary markers: Constraints in instructions.md tell the agent to only use user-provided data and to validate inputs. Capability inventory: The skill has filesystem access to export data using the pandas and openpyxl libraries as defined in SKILL.md and claw.json. Sanitization: No specific sanitization or data escaping mechanisms are implemented in the code to handle potentially malicious content within processed files.
Audit Metadata