remember
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform directory management and create symbolic links, ensuring that documentation and skill configurations are synchronized between platform-specific and tool-agnostic directory layers (e.g., .agents/ and .claude/).
- [PROMPT_INJECTION]: The skill facilitates the persistence of instructions that influence future agent behavior, creating a surface for indirect prompt injection.
- Ingestion points: Arbitrary natural language input is accepted via the 'thing to remember' argument in SKILL.md.
- Boundary markers: No explicit security boundaries or isolation markers are implemented when writing the user-provided content into the target documentation or skill definition files.
- Capability inventory: The skill has broad capabilities to read, write, and edit critical agent instruction files (.agents/skills/*/SKILL.md) and configuration documents (AGENTS.md, CLAUDE.md).
- Sanitization: While the skill rephrases input during 'Intent Distillation', it does not perform formal sanitization or escaping of content before persisting it into the agent's knowledge base.
Audit Metadata