skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/cwicr-change-order/Gen Agent Trust Hub
cwicr-change-order
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data in various formats to generate cost estimates and documentation, creating a surface for potential indirect prompt injection attacks.
- Ingestion points: The assistant accepts project data and files in CSV, Excel, and JSON formats as described in
instructions.md. - Boundary markers: There are no explicit instructions to the agent to treat input data as untrusted or to ignore embedded instructions within that data.
- Capability inventory: The skill has filesystem permissions and uses Python to write Excel documents via
export_change_orderinSKILL.md. - Sanitization: The implementation logic in
SKILL.mdperforms calculations but does not include sanitization or validation of string content within the input data. - [EXTERNAL_DOWNLOADS]: The skill references external resources associated with the vendor,
datadrivenconstruction. - Evidence: The skill points to
https://datadrivenconstruction.ioand the GitHub repositoryhttps://github.com/datadrivenconstruction/OpenConstructionEstimate-DDC-CWICRfor reference and data source information. These are documented neutrally as official vendor resources.
Audit Metadata