careful
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a PreToolUse hook to execute a local validation script (
check-careful.sh) against all bash commands processed by the agent. It also contains a bash snippet for environment setup that creates directories and logs data.- [PERSISTENCE]: The skill establishes a local tracking mechanism by creating a directory at~/.gstack/analyticsand appending session metadata to askill-usage.jsonlfile.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to parse and evaluate untrusted shell commands as its primary input source for safety filtering. - Ingestion points: Command strings passed to the Bash tool as defined in the SKILL.md hook configuration.
- Boundary markers: Not explicitly defined in the provided configuration files.
- Capability inventory: The skill possesses Bash execution and local file system write capabilities.
- Sanitization: Command evaluation and filtering logic is delegated to the
check-careful.shscript.
Audit Metadata