build-domain-ontology
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a purely instructional framework for managing domain knowledge. It lacks any commands for file system modification, privilege escalation, or persistence.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external datasets such as scientific concept records and relation evidence.
- Ingestion points: Data provided via the
domain_scope,concept_records, andrelation_evidencefields in the input contract. - Boundary markers: None identified; the instructions do not specify the use of delimiters or 'ignore' instructions for the ingested content.
- Capability inventory: The skill coordinates the loading of other specific sub-skills (e.g.,
extract-concepts,atomize-concept) but does not directly invoke shell commands, network requests, or dangerous file operations itself. - Sanitization: No specific sanitization or validation logic is defined for the input data.
Audit Metadata