mend-agent-rules
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates a surface for indirect prompt injection by extracting metadata (name and description) from arbitrary repository files and writing it into core agent instruction files ("CLAUDE.md" and "AGENTS.md").
- Ingestion points: Metadata is extracted from "SKILL.md" files located within the ".claude/skills/" directory.
- Boundary markers: The skill does not implement delimiters or specific instructions for the agent to ignore potentially malicious commands within the extracted content.
- Capability inventory: The skill (SKILL.md) instructs the agent to perform file-write operations on the primary configuration files that govern agent behavior in the workspace.
- Sanitization: No logic is provided to sanitize, validate, or escape the content extracted from external files before it is synchronized into the target documentation.
Audit Metadata