skills/ivansamartino/gstack/careful/Gen Agent Trust Hub

careful

Pass

Audited by Gen Agent Trust Hub on Nov 8, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses a PreToolUse hook to execute a local shell script (bin/check-careful.sh) every time the Bash tool is invoked. This script analyzes the proposed command for dangerous patterns.
  • [DATA_EXPOSURE]: The skill collects analytics about its own usage, including the current git repository name and timestamps. This data is stored locally in ~/.gstack/analytics/skill-usage.jsonl and is not transmitted over the network.
  • [SAFE]: The deterministic detection of destructive system commands is a false positive in this context. While the bin/check-careful.sh script contains strings like rm -rf and DROP TABLE, these are used as regex patterns to detect and warn against destructive actions, rather than to execute them.
  • [DYNAMIC_EXECUTION]: The hook script uses a Python fallback to safely parse JSON input and extract command strings when standard shell utilities (grep/sed) fail, avoiding unsafe evaluation of user-supplied strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 8, 2026, 09:19 AM