hooks
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a vendor-provided management script (manage-hooks.sh) located in the plugin installation directory to manage hook states.
- [DATA_EXFILTRATION]: The skill modifies ~/.claude/settings.json to register the itp-hooks. This is a sensitive configuration file, but access is required for the skill's primary function.
- [PROMPT_INJECTION]: User input from $ARGUMENTS is passed directly to a bash command (bash ... $ACTION), creating a potential shell injection surface. Ingestion point: $ARGUMENTS in SKILL.md. Boundary markers: Absent. Capability inventory: Bash tool execution. Sanitization: Absent.
Audit Metadata