vendor-listing
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the use of local Python scripts for validation and ingestion, including
scripts/catalog_verify.py,scripts/catalog_validate.py, andscripts/catalog_ingest.pyexecuted via theuvrunner. - [COMMAND_EXECUTION]: A maintenance script,
drift.sh, is used to synchronize internal documentation fragments within the repository. - [DATA_EXFILTRATION]: The skill mandates strict credential hygiene, requiring that test tokens be read only from environment variables and never written to the filesystem or committed to the version control system.
- [DATA_EXFILTRATION]: A PII scrubbing requirement is included to ensure that identity-related data (such as names, emails, or phone numbers) is removed from provider examples before they are added to the public catalog.
- [PROMPT_INJECTION]: The skill ingests data from external vendor documentation and OpenAPI specifications. While this creates a potential indirect prompt injection surface, the risk is addressed through automated verification scripts and manual data scrubbing requirements.
Audit Metadata