Skip to content
#

execution-authority

Star

Here is 1 public repository matching this topic...

decionis / agent-safe-pipeline

Star 530

Reference architecture for AI agents that propose actions but cannot authorize them — immutable intent capture, an independent Decionis policy verdict (ALLOW/ESCALATE/BLOCK), verified human approval, and a SafeExecutor that consumes a single-use intent-bound grant.

typescript mcp authorization reference-architecture ai-safety human-in-the-loop ai-agents policy-as-code ai-governance agentic-ai execution-authority decionis ai-agent-permissions
  • Updated Aug 31, 2026
  • TypeScript

Add this topic to your repo

To associate your repository with the execution-authority topic, visit your repo's landing page and select "manage topics."

Learn more