Summary
AgentPort launched an open-source gateway that lets autonomous agents connect to external services without direct access to API keys. It adds per-tool approval policies so teams can auto-approve low-risk actions, require approval for sensitive ones, and deny destructive paths entirely.
What changed
AgentPort publicly launched its agent security gateway with prebuilt integrations, approval workflows, and self-hosted deployment docs.
Why it matters
As agents start taking real actions across Stripe, Slack, databases, and developer tools, permissioning becomes part of the product surface, not just infrastructure plumbing. AgentPort turns approval policy into a first-class control layer, which makes agent adoption easier in security-sensitive workflows.
Evidence excerpt
AgentPort describes itself as an open source gateway to securely connect any service to autonomous agents and says agents never see your API keys, with auto-approve, ask for approval, and deny policies.