terminology
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill involves surveying and incorporating information from external sources into a project's terminology index, creating a surface for potential indirect instructions.
- Ingestion points: The skill reads external official documentation, research papers, and competitor product logs as defined in the
references/naming-survey.mdinstructions. - Boundary markers: Curated data is organized into structured markdown tables (Term, Meaning, Elsewhere called, Reference) within
terminology.mdor topic-specific files. - Capability inventory: The skill has the capability to write to
terminology.mdand related documentation files, and it can propose modifications to project instruction files likeAGENTS.mdorCLAUDE.mdto ensure the glossary is used in future tasks. - Sanitization: The
references/naming-survey.mdfile provides explicit instructions to scrutinize AI-generated prose and prioritize primary sources, which helps filter out potentially malicious or low-quality external content.
Audit Metadata