FastClaw
Use case · Agencies

AI agent platform for agencies building bots for clients

Build an agent once, fork it for each client, and deliver it on WeChat, WhatsApp or the web — all from one FastClaw deployment with per-client quotas.

01

Fork agents per client

Clone an agent’s identity, skills and model defaults into a new user’s namespace with forkFrom — the building block for "client buys a bot".

02

Per-client quotas

Each user has an agent quota: unlimited, a fixed number, or zero so only you can provision their agents.

03

Portable configurations

Export an agent’s identity files, prompt mode and skills as a ZIP and import it on another deployment after a preview.

04

Delivered where clients’ customers are

Connect WeChat or WhatsApp by scanning a QR code, or embed the agent behind the OpenAI-compatible API in a client’s app.

05

Hand over without passwords

Mint single-use login links so a client can open their console as their own account.

Build once, deliver many times

Agency work repeats: the same support bot, sales assistant or booking agent, adapted for each client. In FastClaw an agent is a set of Markdown files plus skills and model defaults, so you can design a template agent and fork it into each client's account through the runtime API.

Meet clients on their channels

Many clients need a bot where their customers already are. Connect WeChat or WhatsApp from the agent's channels page by scanning a QR code. For WhatsApp, use a dedicated number: it connects through an unofficial protocol and automated use can get a number banned.

Control costs per client

Route each client's agents to the models that fit their budget, for example through OpenRouter. Usage is reported per account, so you can bill each client for what they use.

Frequently asked questions

Try it on your own agents

Start free in FastClaw Cloud, or self-host the single binary in a minute.