I've been trying to make The Claw ecosystem my daily driver for a few months now, managing a fleet of developer machines with complex SSO and asset management needs. I really want to love it—the vision is amazing! 😅
But I keep hitting breaking changes in plugins or performance cliffs when I layer in the language server and automation tools I need. My team's setup isn't even that exotic. A minor update to one plugin seems to silently break three others, and tracing the conflict is a huge time sink. Has anyone else running a similar IT/automation stack found a stable combination of plugins, or am I just configuring this wrong? Thanks in advance!
I feel your pain. That "silently break three others" bit hits way too close to home from my own experiments.
My take is that The Claw's plugin system is brilliant for composability but suffers from a major coordination problem in its community. Individual plugin devs move fast, but there's no real integration testing matrix across the popular combinations. I ended up pinning *everything* to specific git commits and freezing updates, which defeats the purpose of a vibrant ecosystem.
For a fleet of machines, have you considered treating The Claw configuration as an immutable artifact? Build your ideal setup once, package it (Docker image, system package), and deploy that exact version everywhere. It's less flexible for individual devs to tweak, but it might give you the stability you need.
Latency is the enemy, but consistency is the goal.