@dosco Great. I'll see if I can find some time to take a stab at it.
English
Anthony McClosky
13.3K posts
Anthony McClosky
@amcclosky
Software Developer and sports fan. Currently: https://t.co/e1BQaAQ8iI Previously: @realtordotcom, @OfficialOpcity, @BlackLocus, Optaros, @BaylorECS
Austin, TX Katılım Ekim 2009
4.8K Takip Edilen783 Takipçiler
@amcclosky we have snowflake so big query won’t be that hard it was on the list just couldn’t figure out a good way to add it to the integration test harness same issue with snowflake. in short yes open to contributions
English
@RhysSullivan is the more interesting thing to you the opinionated control flow or that it has a large standard library of first-party packages? something I'm not thinking of?
English
people don't realize error handling is the least interesting thing about effect
Mario Zechner@badlogicgames
effect wouldn't be necessary if TS had checked exceptions. don't @ me.
English
Holy crap this feels like a massive blow to anyone that uses claude -p, which is a lot of the tools out there.
Fortunately, Solo (soloterm.com) uses the real Claude CLI, so we're safe.
ClaudeDevs@ClaudeDevs
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
English
@IanLandsman you were thinking of Codex Spark with is different than /fast mode which is an actual usable version of the fast mode in Claude. Usable in the sense that it pulls from your subscription credits instead of “extra usage” credits like Claude. cc @aarondfrancis
English
@aarondfrancis Been using the soloterm version of this as my go to for days, weeks?, who knows. 🙃
English
@aarondfrancis Oh I was hoping you had worked on something to pick up the agents Claude was spawning on it’s own.
English
@simonw @bradenjhancock I thought this post was interesting and I am interested in fine-tuned “helper” models for larger agent systems. x.com/RampLabs/statu…
Ramp Labs@RampLabs
English
@bradenjhancock Asking because about once every six months I put out a call for examples of fine tuning LLMs working in practice and, aside from the obvious examples like Cursor, I don't often get many great answers
English
Fine-tuning frontier models was a mostly dominated solution anyway. Use an untuned frontier model for prototyping or personal use, but for the large majority of production use cases where you’re going to be doing the same task thousands to millions of times and/or your business depends on the quality of your solution, fine-tuning is basically required. And if you’re fine-tuning anyway, why not get the dramatically better cost, latency, privacy, and control benefits of building on an open source base. The gap between fine-tuning vs not is usually much larger than the gap between closed vs open.
English
@aarondfrancis @dillon_mulroy But they are more ephemeral/ temporary than people think.
English
@aarondfrancis @dillon_mulroy I’ve come around to the idea that an individual tool or product shouldn’t need a skill to work well but as a format for sharing a workflow that compose multiple tools in to something useful that can be shared they are great.
English
i think skills are a mistake and the wrong abstraction. i almost never want my agent auto invoking them and i have built custom tooling to "toggle" them on/off to prevent them from always being present in my context window.
English
@IanLandsman @TokenTownPod I just want y’all to never launch it. Just keep pushing forever.
English
We're pushing @TokenTownPod another week. You will get your entertainment when we are ready to provide it and not a minute before.
English
@romainhuet More hooks to let third-party devs interact with the in-app codex browser
English
We’re thinking about the next wave of Codex plugins.
What’s one you’re missing today?
English
@aarondfrancis @mattpocockuk Eventually I think you should offer some sort of (optional) packaged workflow skill. You could do worse than his as a starting point.
English
@aarondfrancis @mattpocockuk Its a pretty neat bit of DevEx and AgentEx
English
FYI @aarondfrancis the new /setup skill that @mattpocockuk added to his github.com/mattpocock/ski… works nicely with soloterm.com MCP todos and scratchpads.
English
@amcclosky oh you know what, I dont think i tested tables at all! Adding a new todo (in solo (for solo))
English
You ever need a place for your agents to collaborate on some markdown?
Tell it to write a scratchpad. You get full deeplinks, agents can read it, humans can read it.
Make a giant plan, stash it, let multiple agents coordinate off of it.
English