Put it to work

Use your agents from Claude Code, Cursor and VS Code

Put the agents your team has built inside the coding assistants your team already works in.

Captured from the product Demo workspace

Point an MCP client at your workspace and your published agents become tools it can call.

What this changes for your team.

Yekar.AI serves a Model Context Protocol endpoint. A client authenticates with a personal API key and lists your published agents as tools, one per agent, each taking a message and returning the reply with a conversation id it can continue. There is nothing to install and nothing to register. Any client that lets you set a request header can connect - Claude Code, Cursor, VS Code, the Copilot and Codex CLIs; the Claude and ChatGPT apps' own connector screens sign in with OAuth instead and cannot hold an API key yet. The call runs the same published agent as every other surface, so its connections, write approvals and audit trail are unchanged, and the key still acts with its owner's permissions.

How it works in practice.

01
Publish the agent and mint a personal API key under Triggers → API keys.
02
Add the endpoint and that key to your MCP client's server configuration.
03
Call the agent from the assistant, and open the same conversation in Yekar.AI afterwards.

What you can plan around.

The behaviour you can design against, stated concretely.

One tool per agent with a published setup; the client re-reads the tool list rather than caching names.

A call that has not finished within the wait budget answers with the agent's state and its conversation id, not a protocol error, so the assistant can report progress and continue.

The endpoint adds no authority: the same key, the same permission checks, and execution still requires access to the agent's domain.

Authentication is an API key in a request header; there is no OAuth authorization server for this endpoint yet, which is what decides the client list above.

Complete the operating path

View all capabilities

Put it to workStart agents from connected events

Move work from polling and handoffs to event-driven runs across key providers.

Put it to workStart agents through an API trigger

Make a published agent part of your own application or service.

Run it reliablyNever lose track of accepted work

Give every accepted unit of work a durable identity and history before execution begins.

Bring one real process