ClawMetry Agent Builder

Showcase

Research desk

Plans a long research task, splits it across subagents, and comes back hours later with something structured.

For: Anyone whose questions take an afternoon, not a second

What it runs on

Runtime: deepagents — LangChain deep agents with planning, subagents, and skills
Hosting: Standard — $9/month, 1 vCPU / 1 GB micro-VM. First agent month free.
Model: managed model at provider cost · your own API key
Channels: telegramslackdiscordemailmatrixwhatsappsmswebchat

How it works

deepagents runs LangChain's deep-agent loop: an explicit plan, then subagents against it.
Per-chat threads are checkpointed, so a long task survives a restart.
This is the shape where watching cost per conversation matters most — a plan that keeps re-planning is the expensive failure.

The brief

This is what goes into the builder. It is a starting point — the more of your own situation you put in it, the better the agent you get out.

Build me an agent that can take a research question, plan it out, work through it over a few hours and come back with a structured writeup.

Read more first

If you want to understand the runtime before you deploy, Host deepagents covers what it is good at, where it goes wrong, and what it costs to keep running.

Build this agent

Opens the builder with the brief above already written. Change anything before you send it — first agent month is free.

Start with this brief →

← All examples · Guides · Pricing