Back to Building what's next (Interview).
Why Not Let OpenClaw Write the Email Tool?
Zeno Rocha refused to let OpenClaw write the email command line because giving it an open API spec would generate a badly written command line, and he hated that outcome.
Internal command line tools had already shown up in hackathons, yet he had long felt that nobody would open a terminal just to send a transactional email, a password reset, a welcome email, or a usage email. Configuring a domain and reading logs turned out to be the jobs worth doing from the terminal. While he was spinning up OpenClaw instances for the team and for his own use, handing over the open API spec was not an option, so the team built the command line itself. They chose TypeScript so it could be open source. They debated Go, then picked what was better for open source contribution rather than deployment, build, or performance. Shipping the tool took a long time, and shipping the MCP server took a long time too.
After the MCP server was public, clients required an authentication method by default. The team went back through the whole authentication layer, including how every request is saved, so a request could be authenticated in more than one way and not only with an API key.
Building what's next (Interview) ·
From the transcript
What took you so long? Yeah. Were you just too busy? But like, I know in three years time you got so much to do, but like, seriously, why did it take you to this year when the boom happens to make an official CLI? Yeah, that's a great point. That's a great point. We built CLIs over time internally, like during hackathons and stuff like that.
But I've always felt like, man, who's going to want to send an email from the terminal? Like we are trying to help businesses to send, you know, transactional emails, reset password, welcome emails, like whatever you need to send, right? Usage emails. But would someone go and like spin up their terminal and try to send an email from there? My feeling was like, oh, maybe not.