itsol-workflow-mode

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests repository-level configuration from .itsol.md files to resolve workflow modes and cumulative restrictions. This ingestion point allows external data to influence the agent's decision-making process and autonomy levels.
  • Ingestion points: Reads root and project-specific .itsol.md configuration files.
  • Boundary markers: Explicitly instructs the agent to obey platform-level safety rules and provides examples of unambiguous user instructions to prevent accidental autonomy.
  • Capability inventory: Controls the resolution of workflow modes that gate manual planning approvals and influences the transition between subagent execution states.
  • Sanitization: Utilizes a mandatory intersection logic for cumulative restrictions and requires subagents to return a 'blocked' state if authority configurations are inconsistent.
  • [SAFE]: The skill implements a 'Protected Actions' framework that mandates explicit authority for high-risk operations including destructive data actions, production deployments, and secret exposure, ensuring these remain outside the scope of general workflow autonomy.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 02:31 PM
Security Audit — agent-trust-hub — itsol-workflow-mode