The emails go through quickbooks/accounting software, Clawbolt doesn't have any direct email client. Use of tools is on a gradual permission basis like Claude code, and Clawbolt doesn't have any general code access or web access. I think you highlight an important point though that prompt injection continues to be a hazard of AI agent use, though tools continue to be developed to fight against it. The goal is to lock Clawbolt down as much as possible to help users avoid the security hazards of systems like openclaw, but this is definitely something that we'll need to watch and be careful about!
Very cool! Quickbooks is an interesting area because imo it's so important to be super super careful with how that interaction model works. Narrowing on a very specific vertical like small business contractors I think should help to make sure that we can nail the interaction model to make it easy and safe.
Nathan here, MLE at Mozilla.ai. We wrote this because the AI productivity conversation has been almost entirely about desk workers, and we think that's a miss.
My buddy runs a general contracting business. He's good at what he does, but he spends a lot of evenings on a laptop chasing invoices and scheduling follow-ups instead of hanging outwith his family and friends. Most tools being built right now weren't built primarily for him.
Clawbolt is our attempt at a purpose-built AI assistant for people in the trades. It lives in messaging apps they already use, connects to things like QuickBooks and Google Calendar, and is proactive rather than waiting to be asked, just like OpenClaw. Following up on unpaid invoices, flagging material cost changes, that kind of thing. Open source core at github.com/mozilla-ai/clawbolt if you'd like to check it out and give it a star.
We're early and looking for input from people who actually work in the trades or run small trade businesses. Would love to hear what resonates and what we're missing.
Glad to hear it! oooh interesting, I've seen worktrunk but haven't explored it too deeply yet.
The sounds are sort of an easter egg: you can enable/disable them from the settings page (hit 's' in the TUI). You probably are experiencing them being random because the default is to select "random" sounds (see the settings TUI screen). I don't use the sounds much so I'm waiting for a contributor to suggest an improvement.
It's the wonderful part about OSS and 'mission-driven' projects. If the mission is not to make money, then a project is free to reject addons/etc that might be lucrative but not add value to the core of the product
Great idea and name the danger here which I'll be interested to track is how do you keep this "nano"? Since it's built for you, you'll continue adding features i assume which over time will make this not very nano. I guess I'm wondering if there could be some small design tweaks of the repo that make this usable as a long term "fork the base and make it your own" concept
It makes me chuckle every time the state changes lol. But by making it configurable hopefully it can actually be a useful feature for people (fingers crossed)
It’s Sunday night/Monday morning, I just got back from a great day at FOSDEM 2026, and since I’m on the east coast time schedule, might as well tackle some outstanding issues on agent-of-empires.
Step 1, backlog prioritization.
Step 2, realize that truly the most important thing the tool is mising is AoE sound effects.
Step 3, add configurable sound effects.
Now you can optionally enjoy being told “wololo” when Claude needs to you to confirm your plan before executing the implementation. Enjoy.
Go to docs to see how to install the open source sounds (the internet can show you how to grab the real aoe ii sounds, because of copyright concerns I didn’t bake it into the codebase).
Excited to finally be able to give this a try today. I'm documenting my experience using aoe + OpenCode + LM Studio + GLM-4.7 Flash + Mac Mini M4Pro 64GB Mem on this thread if anyone wants to follow along and or give me advice about how badly I'm messing up the settings
I had the same experience using smolagents. Early 2025 it was a competitive approach, but a year later having a small subset (<10) of flexible tools is outperforming the single-tool approach.
Nathan here: Anushri and I wrote this after many discussion in our team here at mozilla.ai. Not only from experiences managing contributions in our own OSS code but also in extenral OSS codebases we contribute to ourselves: we've seen varied methods for handling AI usage when it comes to contributions.
Hopefully this post helps kickstart the discussion about what we should do to help keep OSS codebases a wonderful place to contribute and learn from each other.
Regardless of how you feel about AI generated code, we think it's crucial that we talk and learn from each other. AI Coding is a new world and it's here to stay, learning from each other about how to navigate learning and optimizing our new skill sets is important.
Thanks for giving it a try! If you can think of any improvements I'm certainly all ears. What i've been doing is using the ios + termius + aoe setup to just keep tasks moving forwards. I log in from my phone, see which sessions are waiting for me, and then type just enough to keep them moving forwards until I get back to my desktop where then I can use aoe again to keep them moving
Good questions! All of that is handled inside of the coding agent (Claude Code / OpenCode). `aoe` is responsible for managing the agent sessions: It does keep those sessions persistent in tmux even if you close `aoe`, but if you restart your system, resuming a sessions would rely upon opencode/claude code storing the conversation history for each session, which I believe they do but would need to be manually restored for each agent.
THIS!!! I'm right there with you. Imo, using aoe (or whatever tool you can find for session management) is the key unlock to help give us all the patience to try out the ollama/lmstudio/mlx providers to see how they really perform. We're lacking information about what work is really being done with local coding models, at least somewhat because imo most github open source repos don't ask you to share what LLM you used to help you, and they should.
We need to start sharing our specific success and failure stories from using local llms for coding