gabriel-mode
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill does not contain any detected prompt injections, hardcoded credentials, unauthorized network operations, or persistence mechanisms. All external references are for documentation purposes and target well-known repositories.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes data from external tickets and subagent outputs.
- Ingestion points: Untrusted data enters the context through project tickets and worker handoff evidence during the orchestration phases defined in
SKILL.md. - Boundary markers: The skill references a
handoff contractfor managing data flow, although specific delimiters for external content are not explicitly defined in this file. - Capability inventory: The main agent orchestrates worker subagents and performs file-write operations to a local temporary directory (
<system-temp>/gabriel-mode/) as specified inSKILL.md. - Sanitization: Risks are mitigated by the "Main judgment" and "Review the combined result" steps, which mandate independent verification of all content against established project rules and "Done when" criteria.
Audit Metadata