competitor-profiling
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from external sources, which is a standard surface for indirect prompt injection attacks. \n
- Ingestion points: External URLs, competitor websites, documentation, and community reviews are accessed and processed as evidence (referenced in SKILL.md and references/profile-workflow.md). \n
- Boundary markers: The workflow instructions include separating raw evidence from conclusions and recording source metadata, providing structural separation, but lack explicit instructions to disregard potential commands within source materials. \n
- Capability inventory: The skill uses agent capabilities to perform research, profile generation, and market comparison. \n
- Sanitization: There are no explicit instructions to sanitize or validate the content retrieved from external sources before it is analyzed by the agent.
Audit Metadata