git-helper
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill integrates external data through a template variable, which creates a surface for indirect instructions to influence the agent.
- Ingestion points: The $ARGUMENTS placeholder in SKILL.md allows for the injection of user-supplied content, such as git diffs or commit history, directly into the instruction context.
- Boundary markers: The skill lacks clear delimiters or protective framing (e.g., XML tags or clear 'ignore embedded instructions' directives) around the interpolated data.
- Capability inventory: The skill is configured within an environment where Bash is an available tool, providing a mechanism for command execution if the agent is successfully manipulated.
- Sanitization: No explicit sanitization, filtering, or instruction-following constraints are applied to the external data before it is processed.
Audit Metadata