cancel
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Provides bash scripts for manual state cleanup and process management, including file removal, project root detection, and tmux session termination.
- [REMOTE_CODE_EXECUTION]: References and executes an external Node.js script located within the plugin's root directory to handle process cleanup.
- [DATA_EXFILTRATION]: Accesses local application-specific directories including ~/.claude/teams/ and retrieves repository metadata to derive project identifiers.
- [PROMPT_INJECTION]: Contains instructional directives and processes data from local state files which are then used to drive cleanup logic. This involves ingesting untrusted data from local configuration and state files: 1. Ingestion points: Reads ~/.claude/teams/*/config.json and .omc/state/ JSON files. 2. Boundary markers: Absent. 3. Capability inventory: Includes file system deletion, process management, and script execution across all scripts. 4. Sanitization: None specified for the contents of the ingested state files.
Audit Metadata