... about the Agents dashboard . It already reads what your coding agents load before they start; now it also finds what they leave behind on disk, helps with a skill listing that has grown past its room, and reads Claude

... Lancetta 0.62.0 is out Maintenance now cleans up after your coding agents, and reads Claude

... FinderGit 0.55.0 is out Agents now reads more of what ClaudeCode really loads, and checks more of what it runs. 1. More of what loads The rules of subfolders, your organization's managed rules and the instructions its policy carries, each counted in the launch it joins. Rules are matched from the folders above into the repositories below. 2. No silent gaps A rule behind a link

... Lancetta 0.63.0 is out Maintenance now reads more of what ClaudeCode really loads, and checks more of what it runs. 1. More of what loads The rules of subfolders, your organization's managed rules and the instructions its policy carries, each counted in the launch it joins. Rules are matched from the folders above into the repositories below. 2. No silent gaps A rule behind a link

... agent. So I built PMDesk. It's a macOS app for the person directing the agents, not for writing code by hand. What's in it today Many projects and sessions, no confusion: one tab per project with Claude

... need to make some quick changes in one file. The difference appears when my changes affect six or seven files. When using Cursor, I usually review as I go because the diff is visible in the editor. When using ClaudeCode in the terminal, I provide a larger instruction and then review at the end. The first one feels more secure, while the latter feels faster if I already know what I want. I changed the name of a basic ... ... much can the agent change the

... Product Hunt! AgentDock launched here on 10 October. Since then it has had eleven updates, and the latest one is the biggest: AgentDock now works with many more agents than ClaudeCode and Codex. New in 1.3: more agents Beyond

... logged the raw requests from one ClaudeCode session on a real repo to see where the input tokens actually go. I expected the files the agent read to dominate. Input was about 123k tokens: ~74% tool definitions ~21% conversation, including a 40 KB file the agent read ~5% system prompt So most of it was the list of tools the agent can call, sent again on every request. The file I was worried about was a small slice ... ... with a grain of salt. My guess is it scales with how many MCP servers you have connected, since each one adds its own tool definitions. If you want to check yours without any extra tooling, " /context " inside

... everyone, Said here, the maker of CrewTower. On October 1, ClaudeCode shipped mods. Hooks could already tell an outside app what

... Morse, an indie dev in Thailand. I'm working on AgentValley, a free and open source Mac app that sits in the notch and shows every ClaudeCode session you have running. A crew of pixel programmers waves when an agent is waiting on you, and you can answer from the notch. It isn't launched here yet. I'm waiting for Apple to sign it so it installs like a normal Mac app. Until then I'm mostly here ... ... look around and try other people's launches. What are you working on? Comment from George Basov(@gbasovv): Welcome Morse! This one is aimed right at me. I usually have several

... tech lawyer and ex-founder. I can read code, but beyond HTML and CSS I can't write it. Since my first vibe-coded

... Guys, So I use ClaudeCode a lot for most of the day. Sometimes to implement long plan documents I spend hours refining, or just have it perform tasks that span hunting and fixing bugs all the way to implementing new features. What kept bugging me though, along the way, was where some of the output went during its thinking. I'm a visual person, so I like to see what

... ClaudeCode reads your repo, edits files, runs shell commands and calls MCP tools. That's what makes it useful, and it's also what makes it a security surface no IDE assistant ever was. Its built-in permissions, sandboxing and managed settings reduce the risk meaningfully. They don't remove it. What are your best practices to secure your @

... built a menu bar app that shows what spends your ClaudeCode limit looking for people to try it and tell me what's missing. It started with my own skills folder: I couldn't tell which skills ever fired or what they cost. SkillKeeper reads your local

... ClaudeCode makes it easy to ship interfaces quickly, but builders are often too familiar with their own product to notice where a first-time user gets confused. I m working on HallwayTest , an MCP workflow that helps

... everyone Like a lot of you, I use AI coding agents (Claude

... guys, i switched to codex recently because fable is not anymore included in the plans, and honestly gpt 5.6 Sol surprised me by his efficiency. Which one do you prefer, ClaudeCode or Codex ? Comment from Moss Ab Mirande-Ney(@moss_ab_mirande_ney): @busmark_w_nika   I understand i think we have different usecases :) Comment from Siarhei Hamanovich(@hamanovich): Last 2 months I use Codex more. I switch depending on the job. For a change that touches ... ... walking away wants the opposite. i suspect a lot of the disagreement in threads like this is people using two different products and calling them one. Comment from Nika(@busmark_w_nika): I was more used to

... That was basically the problem I wanted to solve. I really like the Codex app especially how clean and focused the interface feels but for harder coding and architecture tasks I still often want to use Claude

... used to end the day by asking ClaudeCode something like: Can you write my daily update based on what I worked on today? The problem was that

... Hello guys, for the context, in Murmell, we run CLI sessions like Claudecode and Codex, and damn this mf is consuming too much RAM around 800mb. For comparaison Codex consumes about 150mb max since openAI rebuilt it in rust. This is forcing us to keep plans a lil high due to cloud costs, we are thinking into rebuilding our own version of it in rust. Does anyone know about other

... what I could find from benchmarks and my own week of switching between tools, and a few things genuinely surprised me. The efficiency numbers aren't close, and not in the direction I expected. One widely cited test had ClaudeCode completing an identical multi-file feature task using around 33,000 tokens with zero errors , while Cursor Agent burned roughly 188,000 tokens on the same task and hit several errors along ... ... same deliverable. Separately, other community head-to-heads put

... everyone! Launching a new feature of LangWatch today on ProductHunt, developer focussed but not just dev's anymore, so also curious what each of you think. You know what ClaudeCode costs per seat. But do you know which sessions burned the budget, which model did the work, or whether your cache paid off? LangWatch now tracks every

... internet and naturally converged on the same tools. Curious what everyone else is actually using: What's open on your machine today? What did it replace? Where do you still not trust it? Comment from Sabber Ahamed(@sabber_ahamed): ClaudeCode, open pretty much all day at this point. replaced Cursor for me for anything past a single-file tweak, mostly because I can describe the outcome and walk away instead of babysitting a diff. still keep Cursor around ... ... quick visual CSS stuff where seeing it live matters more than describing it. Comment from Tarqiya Forgah(@tarqiya_forgah): I've consistently used Codex to create frontend developments for web applications. To be honest, I would prefer using

... been using Codex, ClaudeCode, and OpenClaw for different kinds of work: Codex (now @ChatGPT by OpenAI )is strong at long-context planning and execution. @

... ClaudeCode is amazing. Agentic workflows just... work - at least better than last years options for vibe