opc-niche
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze content from untrusted external sources, including community forums (Reddit, Discord, Slack), software review platforms (G2, Capterra, App Store), and search engine result pages (SERPs) as outlined in
references/signals.md. This represents a surface for indirect prompt injection where malicious instructions could be embedded in the data being researched. - Ingestion points: External community posts, competitor reviews, and search results described in
references/signals.mdandSKILL.md. - Boundary markers: The instructions lack explicit delimiters or 'ignore embedded instructions' warnings for the data being processed.
- Capability inventory: The skill utilizes network search and browsing capabilities to fetch the external data.
- Sanitization: There is no mention of sanitization, escaping, or filtering of the external content before it is processed by the agent.
Audit Metadata