skills/lev-os/agents/work-mvp/Gen Agent Trust Hub

work-mvp

Pass

Audited by Gen Agent Trust Hub on Mar 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill dynamically executes commands retrieved from FlowMind sessions using the 'lev exec' utility. This is the primary intended mechanism for task orchestration within the Lev ecosystem.
  • [PROMPT_INJECTION]: The skill possesses an indirect instruction ingestion surface where steps are fetched from an external session state. 1. Ingestion points: 'lev flowmind session get' in SKILL.md. 2. Boundary markers: No explicit delimiters are present in the execution loop. 3. Capability inventory: Can execute commands via 'lev exec' and manage beads via 'bd'. 4. Sanitization: No content validation is performed. This is consistent with the skill's design as an orchestration agent.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the '@lev-os/flowmind' Node.js package, which is a resource owned and maintained by the author 'lev-os'.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 21, 2026, 06:49 PM
Security Audit — agent-trust-hub — work-mvp