Skills
skills/popup-studio-ai/bkit-claude-code/control/Gen Agent Trust Hub

control

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to manage local state files and execute internal audit logging scripts. These operations are restricted to the local filesystem and the project's own modules within the .bkit/ and lib/ directories.
  • [DATA_EXFILTRATION]: No network activity or external data transmission was identified. All operations are confined to the project's local directory structure, interacting exclusively with internal runtime and state JSON files such as .bkit/runtime/control-state.json and .bkit/state/trust-score.json.
  • [PROMPT_INJECTION]: The instructions do not contain bypass patterns or attempts to override safety guidelines. Automation level increases (escalation) are protected by a manual approval gate requiring explicit user confirmation via the AskUserQuestion tool.
  • [REMOTE_CODE_EXECUTION]: The skill does not download or execute code from external sources. It relies exclusively on local, project-specific modules such as lib/control/automation-controller.js and lib/audit/audit-logger.js for its logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 10:30 AM