agent-md-updator
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to read and modify sensitive configuration files in the user's home directory, specifically
~/.claude/CLAUDE.md,~/.agents/AGENTS.md, and~/.gemini/GEMINI.md. These files typically contain system instructions that define the agent's behavior. - [COMMAND_EXECUTION]: The workflow directs the agent to execute filesystem operations, such as checking symbolic link relationships and creating backup copies of system files in
~/tmpbefore performing modifications. - [COMMAND_EXECUTION]: By providing a mechanism for the agent to alter its own persistent instruction set, the skill creates a potential vector for indirect prompt injection or persistence, where malicious instructions could be permanently added to the agent's core configuration.
Audit Metadata