qbr-facilitator
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [No Executable Code] (SAFE): The analyzed files (
rules/_sections.md,rules/preparation-executive.md,rules/program-design.md) are documentation and business process templates. There are no script files (.py, .js, .sh), executables, or configuration files that initiate code execution. - [Indirect Prompt Injection Surface] (LOW): The skill defines a workflow that involves processing untrusted external data.
- Ingestion points:
rules/preparation-executive.mdsuggests gathering 'Company Intelligence' (press releases, news) and 'Stakeholder Research' (LinkedIn posts), as well as 'Customer Input Request' (emails). - Boundary markers: Absent from the provided templates.
- Capability inventory: None. The skill does not contain logic for file writing, network operations, or subprocess execution.
- Sanitization: Absent. Since the skill is purely instructional, it does not provide methods for filtering or escaping external content.
- [Data Exposure] (SAFE): The templates use placeholders for business metrics (e.g.,
[ARR],$XXX,[Health Score]). There are no hardcoded credentials, API keys, or automated data exfiltration patterns present.
Audit Metadata