write-outreach-outreach-qa
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill contains no code or executable scripts. It provides a set of criteria and rules for evaluating outreach content, such as factual accuracy, compliance with regional regulations (CASL, AU-NZ), and craft standards (word count, touch frequency).
- [SAFE]: No remote downloads, package installations, or dynamic execution patterns were detected. The instructions are purely textual and operate within the context of the agent's logic for content review.
- [INDIRECT_PROMPT_INJECTION]: The skill is intended to process external data (outreach packs and send files). While this introduces a theoretical surface for indirect prompt injection, the risk is negligible as the skill does not grant the agent any sensitive capabilities (like file system writes or network access) to exploit.
- Ingestion points: Outreach packs and send files processed by the agent.
- Boundary markers: None specified in these instructions.
- Capability inventory: No file-write, network, or subprocess capabilities are utilized.
- Sanitization: None specified; the skill relies on the agent's existing safety guardrails.
Audit Metadata