skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/lien-waiver-tracker/Gen Agent Trust Hub
lien-waiver-tracker
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection as it ingests untrusted data from external files (CSV, Excel, JSON) mentioned in instructions.md.
- Ingestion points: Data is ingested via common formats (CSV, Excel, JSON) as specified in instructions.md.
- Boundary markers: The prompt includes a constraint to "Only use data provided by the user or referenced in the skill."
- Capability inventory: The LienWaiverTracker class in SKILL.md performs only in-memory logic and reporting; no network or subprocess capabilities are present.
- Sanitization: The instructions explicitly require the agent to "Validate inputs before processing."- [SAFE]: The Python implementation is transparent and follows best practices for data modeling. It does not contain any obfuscation, hardcoded secrets, or remote code execution patterns.
Audit Metadata