Everything that shipped in OpenRig since mid-September, in one post: 30 releases, ~2,100 merged PRs, 40+ new features and 20+ improvements. Each one says what it does for you and who built it.
A lot of you are in there. Thank you.
https://t.co/1oqzBiKHBT
btw - did you know you can hand work to your OpenRig team from Claude Desktop, T3 Code, Cursor, etc, or give Grokbot its own seat on the team?
All 32 integrations, what each can do and how to start: https://t.co/UNxxCTmRQU
i used to wonder what terminal sprawl did to the brain. newish addition for me is this but with browser testing.
I'm often watching agents working in separate chrome profiles testing user stories. a few in parallel has been fine, now its dozens at once. overwhelm has a new flavor.
@pbakaus we hit the same thing with openrig. it has an optional typing guard. in draft aware mode an agent's message is held instead of typing over your draft, the sender is told, and nothing is auto delivered later. you read the held ones yourself
been quiet on here. openrig was scaling and i went all in on this video, even shut down my build vm because it kept glitching my audio
it's how i do context for my agents. pretty heady, no idea if it lands, but i'm really excited about it
https://t.co/3KPFzr1J5N
@bradvin really cool brad. streaming each agent's terminal from the daemon into one window is exactly the kind of thing i hoped people would build on openrig. love seeing a community client show up
@daryllnets if you already live in tmux, openrig might fit. it keeps your agents as plain tmux sessions, shows which ones are waiting on you, and lets one agent type a message straight into another's terminal, so reports land in its input. runs on linux too. i'm biased since i make it
@YMiliaresis thanks yiannis. our growth team is a rig too, with a scout reading x and a lead handing out the work. nice idea pulling the roles from agency agents
back to building openrig after about a week making this new video. this was a fun one to make, sprinkled in some cinematic elements.
https://t.co/mwLXv6ocEf
@kerororo66 thanks. that agent wasn't good on day one. it did the same one job for weeks and kept a notes file of what it learned, and that file is what the new agent started from. if you post where your homelab team gets stuck i'll take a look
One of my claude agents has been reviewing and merging openrig's prs in a single session for 25 days, almost non-stop. that session file is 8.7 gb now. aiming to hand the job to a new agent today. inertia...
@elitemaks a mac mini on my desk and a couple of small vms. my agents on different machines message each other through openrig, so i don't ssh around to check on them. one 12 gb vm ran my pr review agent for 25 days straight in a single session
@Quantradin@buildwithneej both review everything, independently. same reviewer role on two models, one on claude code and one on codex, because in my experience they catch different kinds of mistakes
@staskulesh@horosin_ glad the kraków meetup put it on your radar stas, and #92 to #48 in a day is fun to watch. if anything trips you up, reply here or open a github issue and i'll take a look
@dgalarza same three harnesses here, opened in herdr through openrig. one thing i'd add on review, let one reviewer stick around. mine reviewed and merged openrig's prs for 25 days in one claude session before i handed the job over, and its memory of what broke was the part worth keeping
v0.6.12 is about to release. Its a big catchup and has 200+ improvements, mostly from community contributors.
That worldbuilding context engineering was a beast to write 😅
Making solid progress working thru this stack of issues and PR's that built up the last week while I worked on it. Thanks to everyone for this continued stream of ideas and bug fixes!