issue-flow
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external defect reports, which may contain untrusted data meant to influence agent behavior during reproduction.
- Ingestion points: Issue reports, context, and identifiers collected during the 'Intake and acceptance target' phase.
- Boundary markers: The skill employs a structured state machine (intake → reproduce → baseline, etc.) which provides a framework for isolation, though explicit delimiters for external content are not specified.
- Capability inventory: The agent utilizes tools such as
agent-browser,git-workflow, and local reproduction environments to execute code and interact with repositories. - Sanitization: The instructions do not explicitly describe sanitization or escaping mechanisms for data extracted from issue reports before it is used in reproduction steps or testing.
Audit Metadata