tuning-incremental-sync-config
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted metadata from external data sources which could theoretically contain malicious instructions designed to influence agent behavior.
- Ingestion points: The workflow in
SKILL.md(Step 1 and Step 2) utilizesexternal-data-schemas-retrieveandexternal-data-schemas-incremental-fields-createto read schema definitions, column names, and error messages from user-connected databases. - Boundary markers: The instructions lack explicit boundary markers or delimiters to differentiate between external source data and the agent's core instructions.
- Capability inventory: The skill has access to impactful tools including
external-data-schemas-partial-update,external-data-schemas-resync, andexternal-data-schemas-delete-datawhich can modify, wipe, or re-import data. - Sanitization: There is no documented validation or sanitization process for the metadata retrieved from the external source before it is used by the agent to make configuration decisions.
Audit Metadata