skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/data-source-audit/Gen Agent Trust Hub
data-source-audit
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data which could potentially contain malicious instructions aimed at the agent.
- Ingestion points: Data enters the skill via
discover_sources_from_survey(survey responses) andassess_source_quality(pandas DataFrames) as defined inSKILL.md. - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the data processing logic.
- Capability inventory: The skill possesses filesystem write capabilities, as demonstrated by the
catalog.to_excel()method in the Quick Start section ofSKILL.md. - Sanitization: The implementation lack specific sanitization or filtering of external content before processing, relying on basic type conversion for quality scoring.
Audit Metadata