jumploops

8 ms·

121.
Subagents now available in Codex(developers.openai.com)
1 points by jumploops 7mo ago|0 comments
122.
by jumploops 7mo ago
So much of practical CS is abiding by standards created by solo programmers in the past. My university frowned on any industry-related classes (i.e. teaching software engineering tools vs. theoretical CS), but I was fortunate enough to know
123.
by jumploops 7mo ago
Yeah the old adage "what you put in is what you get out" is highly relevant here. Admittedly I'm knowledgable in most of the domains I use LLMs for, but even so, my prompts are much longer now than they used to be. LLMs are t
124.
by jumploops 7mo ago
After "fully vibecoding" (i.e. I don't read the code) a few projects, the important aspect of this isn't so much the different agents, but the development process. Ironically, it resembles waterfall much more so than agi
125.
by jumploops 7mo ago
Yeah exactly, "right way" is in quotes because there is no right way. The most important thing is shipping/getting feedback, everything else is theatre at best, or a project-killing distraction at worst. As a concrete example
126.
by jumploops 7mo ago
This is similar to how I use LLMs (architect/plan -> implement -> debug/review), but after getting bit a few times, I have a few extra things in my process: The main difference between my workflow and the authors, is that I
127.
by jumploops 7mo ago
A lot of these resonate with me, particularly the mental fatigue. It feels like normal coding forced me to slow my brain down, whereas now my mind is the limit. For context, I started an experiment to rebuild a previous project entirely wit
128.
gstack – Garry Tan's Claude Code Setup(github.com)
15 points by jumploops 7mo ago|15 comments
129.
by jumploops 7mo ago
> we launched both a skill and MCP server. My guess is that the MCP was easy enough to add, and some tools only support MCP. Personal opinion: MCP is just codified context pollution.
130.
by jumploops 7mo ago
This is great, giving agents access to logs (dev or prod) tightens the debug flow substantially. With that said, I often find myself leaning on the debug flow for non-errors e.g. UI/UX regressions that the models are still bad at visua
131.
by jumploops 7mo ago
I do something similar, but across three doc types: design, plan, and debug Design works similar to your project.md file, but on a per feature request. I also explicitly ask it to outline open questions/unknowns. Once the design doc (i
132.
by jumploops 7mo ago
I've been experimenting with a few ways to keep the "historical context" of the codebase relevant to future agent sessions. First, I tried using simple inline comments, but the agents happily (and silently) removed them, even
133.
by jumploops 8mo ago
In the context of traditional SaaS, using dynamic secrets loaded at runtime (KMS+Dynamo, etc.). For agentic tools and pure agents, a proxy is the safest approach. The agent can even think it has a real API key, but said key is worthless out
134.
by jumploops 8mo ago
Yes, exactly. The LLM is onboarding to your codebase with each context window, all it knows is what it’s seen already.
135.
by jumploops 8mo ago
Yeah to be clear it will have the same issues as a flyby contributor if prompted to. Meaning if you ask it “handle this new condition” it will happily throw in a hacky conditional and get the job done. I’ve found the most success in having
136.
by jumploops 9mo ago
I’ve found that LLMs seem to work better on LLM-generated codebases. Commercial codebases, especially private internal ones, are often messy. It seems this is mostly due to the iterative nature of development in response to customer demands
137.
by jumploops 9mo ago
I’ve been “testing” LLM willingness to explore novel ideas/hypotheses for a few random topics[0]. The earlier LLMs were interesting, in that their sycophantic nature eagerly agreed, often lacking criticality. After reducing said sycoph
138.
by jumploops 9mo ago
> For complex tasks, Kimi K2.5 can self-direct an agent swarm with up to 100 sub-agents, executing parallel workflows across up to 1,500 tool calls. > K2.5 Agent Swarm improves performance on complex tasks through parallel, specialize
139.
by jumploops 9mo ago
> People have said that software engineering at large tech companies resembles "plumbing" > AI code [..] may also free up a space for engineers seeking to restore a genuine sense of craft and creative expression This resonat
140.
by jumploops 9mo ago
That’s what I used to think, before chatting with the OAI team. The docs are a bit misleading/opaque, but essentially reasoning persists for multiple sequential assistant turns, but is discarded upon the next user turn[0]. The diagram
141.
by jumploops 9mo ago
I think the delta may be an overloaded use of "turn"? The Responses API does preserve reasoning across multiple "agent turns", but doesn't appear to across multiple "user turns" (as of November, at least).
142.
by jumploops 9mo ago
Maybe it's changed, but this is certainly how it was back in November. I would see my context window jump in size, after each user turn (i.e. from 70 to 85% remaining). Built a tool to analyze the requests, and sure enough the reasonin
143.
by jumploops 9mo ago
One thing that surprised me when diving into the Codex internals was that the reasoning tokens persist during the agent tool call loop, but are discarded after every user turn. This helps preserve context over many turns, but it can also me
144.
by jumploops 9mo ago
I believe Claude Code recently turned on max reasoning for all requests. Previously you’d have to set it manually or use the word “ultrathink”
145.
by jumploops 9mo ago
Boot is a misleading term, but you can resume snapshotted VMs in single digit ms (and without unikernels, though they certainly help)
146.
by jumploops 9mo ago
I ran the Gas Town intro post through ChatGPT 5.2 Pro[0] Based on my initial read, and a pass at this summary, it seems mostly right. YMMV Did some further dives into the little public usage data from Gas Town, and found that most of the &q
147.
Claude Cowork runs Linux VM via Apple virtualization framework(gist.github.com)
120 points by jumploops 9mo ago|46 comments
148.
by jumploops 9mo ago
I’ve found that experienced devs use agentic coding in a more “hands-on” way than beginners and pure vibe-coders. Vibecoders are the best because they push the models in humorous and unexpected ways. Junior devs are like “I automated the de
149.
LMArena is a cancer on AI(surgehq.ai)
246 points by jumploops 9mo ago|100 comments
150.
by jumploops 9mo ago
To be fair, the author says: "Do not use Gas Town." I started "fully vibecoding" 6 months ago, on a side-project, just to see if it was possible. It was painful. The models kept breaking existing functionality, overcompl

More ›