maestro-session-manage

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to execute local CLI commands (maestro session, maestro knowledge). These commands are used to query session status, perform lifecycle transitions (complete, archive, unarchive), and manage knowledge candidates. This behavior is consistent with the skill's stated purpose of being a management surface.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external command outputs, specifically JSON responses from maestro session status and maestro knowledge review. While these represent potential ingestion points for untrusted data, the skill uses structured --json flags and strictly defined logic to process the receipts, and it requires manual user intervention for high-impact actions like knowledge promotion.
  • [SAFE]: Several security best practices are observed: the use of revision-fenced transitions (--expected-orchestration-revision) to prevent race conditions, reliance on authoritative CLI tools rather than manual file parsing for critical logic, and mandatory user input steps for knowledge governance decisions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 12:15 PM
Security Audit — agent-trust-hub — maestro-session-manage