Launching today
OLO ToolGate
Self-hosted authorization for AI agents and MCP tools
2 followers
Self-hosted authorization for AI agents and MCP tools
2 followers
OLO ToolGate is an Apache-2.0, self-hosted security gateway for AI agents, MCP servers and tool execution. Apply ALLOW, ASK or BLOCK policies, independent human approvals and single-use execution permits, with audit trails linking identities, decisions and outcomes. Now in beta, with a Docker Compose quickstart for local evaluation.
Free
Launch Team
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Hi Product Hunt, I’m Rahul, creator of OLO ToolGate.
I built it to answer a practical question: should this AI agent be allowed to make this tool call, for this person, right now? ToolGate puts that authorization decision at the execution boundary, with ALLOW / ASK / BLOCK policies, independent approvals and single-use permits.
It is self-hosted and Apache-2.0 licensed. You can govern existing MCP servers or package custom tools, then trace requests, decisions and outcomes.
This is a beta. Tool Host container/network isolation and durable result retry are unfinished; the quickstart is for single-node evaluation. The README documents current limits.
ToolGate is the first current product under OLO (Open LLM Orchestration). OLO Orgs will maintain the open-source work, and I contribute on weekends.
I’d love feedback on setup, trust boundaries and approval workflows. Please report defects at https://github.com/olo-labs/olo-toolgate/issues and follow https://github.com/olo-labs/olo-toolgate/blob/main/SECURITY.md for vulnerability reports. Contributions, examples and documentation fixes are welcome.