fusion-rule-author
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted repository files to extract coding conventions, which could contain malicious instructions to influence the agent.
- Ingestion points: The agent is instructed to scan project files including
README.md,CONTRIBUTING.md,package.json, and existing AI instructions (defined inSKILL.mdStep 2). - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' directives to isolate the content of the scanned files from the agent's instructions.
- Capability inventory: The agent has the capability to read local files and write new rule files to the repository (
SKILL.mdStep 7). - Sanitization: No validation or sanitization of the content extracted from repository files is mandated.
- Mitigation: The risk is significantly reduced by a mandatory review phase (
SKILL.mdStep 6) where the developer must approve all drafted rule files before they are written to disk.
Audit Metadata