cold-outreach-sequence
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is designed for standard marketing and sales operations.
- [SAFE]: Any external URLs, such as brianrwagner.com, are vendor-owned resources that match the author's identity and show no signs of typosquatting.
- [INDIRECT_PROMPT_INJECTION]: The skill uses web_search to ingest untrusted data for outreach personalization. This creates a surface for indirect prompt injection if external web content contains malicious instructions. However, the risk is minimal as the instructions constrain the agent to extracting specific data points (funding, posts) for drafting messages, rather than following commands found in the text. This ingestion is a core requirement of the skill's primary functionality. 1. Ingestion points: web_search calls in SKILL.md and SKILL-OC.md. 2. Boundary markers: Absent. 3. Capability inventory: Limited to text generation and further search calls. 4. Sanitization: Absent.
Audit Metadata