optise-helix-bluf-writer
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data, specifically buyer prompts, proof points, and voice samples, which are subsequently used in the generation of marketing answer blocks. This creates a surface for indirect prompt injection.
- Ingestion points: Data ingestion occurs in SKILL.md Section 4, Step 1.
- Boundary markers: There are no explicit delimiters or boundary markers defined for the input fields in the workflow.
- Capability inventory: The agent has access to web_fetch and web_search tools as noted in Section 0 of SKILL.md.
- Sanitization: The skill mitigates risks through strict structural constraints, including a hard word count limit of 40-60 words and a requirement to strictly follow one of six predefined answer patterns.
Audit Metadata