competitor-monitoring-system
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external websites, which serves as an attack surface for indirect prompt injection attacks. An attacker could place malicious instructions on a competitor's blog or social media profile to influence the agent's analysis or report output.
- Ingestion points: The system uses multiple scrapers (blog, social media, Reddit, HN, reviews) to pull data into the agent's context as defined in SKILL.md.
- Boundary markers: There are no instructions for the agent to treat this external data as untrusted or to use specific delimiters to separate it from its primary instructions.
- Capability inventory: The skill uses network-accessing tools (scrapers) and has file-writing capabilities to save reports to the clients/ directory.
- Sanitization: The playbook does not define any sanitization or validation steps for the scraped content before it is analyzed.
Audit Metadata