unlink-agentic-grimoire-custom
Pass
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local bash script (
unlink.sh) to perform file system operations. - [SAFE]: The script implements a strict denylist (e.g.,
.claude.json,settings.json,projects/) to ensure that sensitive account data, settings, and personal history are never modified or deleted. - [SAFE]: The logic includes a validation step that only removes symbolic links if they point to the exact expected 'root' target, preventing the accidental removal of user-created files or third-party links.
- [SAFE]: The script performs no network requests and does not download or execute code from external sources.
Audit Metadata