investigate

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill preamble and hooks execute several scripts from the gstack toolkit (e.g., gstack-update-check, gstack-config, gstack-telemetry-log, check-freeze.sh). These are part of the skill's own environment and are used for configuration and session management.
  • [DATA_EXFILTRATION]: The skill logs telemetry data, including the current repository name and execution duration, using gstack-telemetry-log. This behavior is gated by a user consent prompt (AskUserQuestion) where the user can opt-out or choose anonymous mode.
  • [EXTERNAL_DOWNLOADS]: The skill performs update checks via gstack-update-check and provides an inline upgrade flow. It also requests user permission to open an educational URL (https://garryslist.org/posts/boil-the-ocean) in the system browser.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 05:34 AM
Security Audit — agent-trust-hub — investigate