consult-bug-ticket
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements an absolute human-in-the-loop gate (Step 5) that requires explicit user approval of all content (GitHub issue body, CRM task, and Gmail draft) and targets before any write operations occur.
- [SAFE]: Explicit PII (Personally Identifiable Information) protection rules are enforced, ensuring that sensitive contact information and internal links are restricted to private CRM systems, while public GitHub issues only contain company names and technical symptoms.
- [SAFE]: The skill identifies and manages indirect prompt injection risks by ingesting untrusted data (emails/pasted text) into a structural template (
issue-template.md) and requiring human review to verify the parsed output before it is used. - [SAFE]: Sensitive business logic, such as client-to-repository routing, is stored in a local configuration file (
bug-routing-config.md) rather than being hardcoded or stored in the skill's source, ensuring confidentiality. - [SAFE]: The skill includes data integrity measures, such as escaping single quotes when updating CRM notes and using 'confirm=true' flags for tool calls to prevent accidental data overwrites.
Audit Metadata