TypeSafe AI for AI agents.
TypeSafe AI integration for AI agents with secure authentication and server-side credential injection. Open Connector runs the OAuth, seals the token in an encrypted vault, and serves TypeSafe AI tools to your agent over MCP or a typed API — credentials injected server-side, every call audited, nothing leaving your infrastructure. Open source (AGPL-3.0) and self-hostable.
# Add your Open Connector MCP server (per-user mcp_url)
$ claude mcp add --transport http \
open-connector <your mcp_url>
# Then ask: “open an issue for the launch checklist”// .cursor/mcp.json
{
"mcpServers": {
"open-connector": { "url": "<your mcp_url>" }
}
}const composio = new Composio({
apiKey: "oc_…",
baseURL: "https://connect.acme.com", // your host
});
await composio.tools.proxy({ … });{
"tool": "TYPESAFE_AI_EVALUATE",
"arguments": { … }
}Real TypeSafe AI actions, managed and audited.
Your user connects TypeSafe AI once; your agent can then evaluate a supplied state against typed questions and return answer probabilities — scoped to the OAuth permissions you grant and the tool allowlist you configure. Every action is least-privilege and written to a tamper-evident audit trail.
- 01Connect
Your user grants TypeSafe AI access once (OAuth) — the token lands in the vault.
- 02Call
Your agent calls a tool over MCP or the typed API; Open Connector injects the credential server-side.
- 03Audit
Every routed call appends a hash-chained audit record — nothing leaves your infra.
Supported TypeSafe AI tools.
Browse 1 tools in the TypeSafe AI catalog. Each action uses your configured authentication and tool allowlist.
Showing 1 tools.
- Evaluate typed questions
- TYPESAFE_AI_EVALUATE
- Ask TypeSafe's System One model to judge a supplied state with named yes/no, choice, or score questions. Returns typed answers and probabilities; check factual conclusions against the original source.
- aiclassificationevaluation
TypeSafe AI integration, answered
- How do AI agents use TypeSafe AI through Open Connector?
- Your user connects TypeSafe AI once with one of its cataloged authentication methods. Open Connector stores the credential in an encrypted vault and exposes TypeSafe AI tools to your agent over MCP or a typed API, with credentials injected server-side on each call.
- Is this a TypeSafe AI MCP server?
- Yes. Open Connector can serve TypeSafe AI as a named MCP server with a scoped allowlist and a per-user connection URL, so any MCP client can call TypeSafe AI actions with credentials injected server-side.
- Where do TypeSafe AI credentials live?
- In your own infrastructure. Open Connector keeps credentials in its own vault and injects them at call time, so they never leave your environment.
Give your agents TypeSafe AI — keep the keys.
Open source, self-hostable, with TypeSafe AI credentials that never leave your infrastructure. Run it from source today.