Runtime security for AI agents

Let agents work.
Protect your data.

Stop data exfiltration from AI agents.

Request a demo

Bullfrog enforces your policy in real time, blocking unauthorized traffic before it leaves the machine.

Outbound requests

<your-machine>

Install a package

GET registry.npmjs.org/express
npm Allowed

Update your repository

POST github.com/<your-org>/<your-repo>/…
git Allowed

Open a pull request

POST api.github.com/repos/<your-org>/<your-repo>/pulls
curl Allowed

Create an issue via MCP

POST api.githubcopilot.com/mcpMCP tools/call · create_issue
Claude Code Blocked

Send data to an attacker’s repository

POST api.github.com/repos/<attacker>/<repo>/issues
curl Blocked

Publish a package

PUT registry.npmjs.org/<package>
npm Blocked

Manage centrally. Enforce locally.

Set policies for the machines running your AI agents from one console. Share policies across machines or define different boundaries, and review outbound activity and policy decisions in one place.

Central console
Example machines, policies and illustrative last activity.
MachinePolicyLast activity
agent-dev-01
 Engineering 2m ago
agent-dev-02
 Engineering 8m ago
research-01
 Research 1h ago
Policy enforced on each machine

Built for the way agents run

Per request, not per domain

Decide on the method, path and body of each HTTPS request, so shared services like GitHub stay usable without becoming a way out.

No CA, no proxy

No interception certificates to create, distribute or rotate, and no custom trust store to configure per tool. Certificate pinning stays intact on supported tools and TLS stacks.

A policy per machine

Give machines different boundaries, or deliberately share a policy. Allowing a site for one machine does not have to open it for the others.

Covers container workloads

Inspect supported container traffic from the host, with nothing baked into your images. Coverage depends on the tools and TLS stacks inside.

Knows which process asked

See process context alongside outbound activity and policy decisions, not just an IP and a port.

Payloads stay local

Requests are inspected on your machine, not sent elsewhere for inspection. Activity metadata is reported centrally, but there is no separate inspection service in the traffic path.

Simple to install. Simple to manage.

  • No proxy
  • No custom CA
  1. 1

    Install

    Deploy Bullfrog on the machines running your AI agents.

  2. 2

    Set policy

    Define policies for outbound requests and remote MCP tools in one central console.

  3. 3

    Govern

    See outbound activity and policy decisions across your machines from the same console.