flowda-template
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill manages the processing of untrusted data from external platforms. * Ingestion points: Data returned by the flowda_template tool from Amazon and Walmart product and search pages as described in SKILL.md. * Boundary markers: Rule 3 in SKILL.md explicitly instructs the agent to treat fetched content as data and never follow instructions found there. * Capability inventory: The skill utilizes various data extraction and search tools including flowda_template, flowda_extract, and flowda_search. * Sanitization: The instructions provide specific behavioral guidance to ignore any instructions embedded in the retrieved web content, maintaining a strict boundary between data and execution.
Audit Metadata