... agents running in parallel... Each one on its own branch. Then comes the merge hehe Conflicts, stale branches

... Codex and Grok for now. Tokens from the other agents get their own series in the hover card and the Tokens widget. The full list, and what AgentDock reads for each agent: agentdockapp.com/agents You asked, it shipped Git branches and worktrees. On launch day someone asked whether the dots can tell apart several Claude Code sessions on one repository. Every agent now shows its branch

... octopus in the stranger-things theme Also in 0.36.0: pressing r now resets the automatic refresh the next one counts from yours and inherits any rate-limit back-off yours ran into. And the supply-chain scan stops mixing branches: a trigger only counts where GitHub can actually start it, and only alongside secrets on the same branch

... empty state needs handling. Errors need messages. Settings need a place to live. None of these feel optional in the moment, but a week later I look up and I've spent most of my time on the branches instead of the trunk. So my question is about where you draw the line while you're still in the middle of building. How complete does the core feature need to be before you call it done ... ... ugly errors, no settings and fix it only if users complain? Or is there a minimum level of finish below which people won't even try it? And more practically: how do you tell the difference between a side branch

... none of it was a config mistake. Our generator page had a link meaning save this to my account, and it was computed from whether you were signed in, so a signed-in visitor skipped the signup step. One branch, on the highest traffic page we have, and it gives the page two bodies. A page with two bodies cannot be handed to a shared cache, so the framework did the correct thing and marked it uncacheable, silently, for months ... ... forwards a visitor who already has one. The page went from two bodies to one, and became cacheable as a side effect. Nine pages moved. One deliberately did not, and that is the more interesting half. Our pricing page branches

... what an agent session actually risks. It edits files in a working tree. If that tree was clean when it started, the entire session's damage is one diff you can read and one reset you can throw away. Branches cost almost nothing, so every task can live on its own one, and merging stays a human decision made after reading the evidence. Underneath that sits the layer most people only meet in a crisis: the reflog. Every commit, reset ... ... rebase, and branch

... nothing, the reviewer's normal experience becomes an empty screen, and empty screens train people to expect empty. Whether the one that is not empty gets read properly on day forty is an open question. It still beats the alternative, because a difference against a background of silence is about as salient as anything gets. And yes to the small pile. Most of what gets called judgement is work nobody has bothered to make mechanical yet. Comment from Aadil Ghani ... ... because pointing at a weakly related ticket becomes a discretionary act that leaves a trace rather than a required one everybody performs. The incentive relocates somewhere auditable instead of disappearing. On ours the honest read is worse than either branch

... landed on: Set aside time each week to specifically work on paying down tech-debt and refactoring. Leverage AI to go deeper when tech-debt surfaces. This starts usually with an audit for a specific topic and then branches into more thorough sub-plans. Acceptance. I'm banking on the number of feature improvements to outweigh the problems a mountain of debt will create. I guess this is the closest to the honest trade-off great product and tech divisions ... ... Fair point and yes, I believe continuous refactoring tilts the equation here. Same can be said about error resolution. As long as instrumentation is in place to catch issues quickly one could say the trade-off of moving fast vs

... actual insight, and it's bigger than it looks: making the notification say why it's waiting is what turns a ping into a triage. "Done" makes you tab back to find out what "done" means. "Needs approval vs finished vs

... once. We moved the PTY off Tauri's IPC channel onto a loopback socket, so now ten agents start near-instantly. Windows got its own overhaul in 0.25.0. The "Not Responding" freezes when you opened several projects or switched branches are gone, because every git and filesystem call now runs off the UI thread. Large files open without the hitch too: the viewer only renders what's on screen and highlights syntax in the background. Git without leaving the terminal ... ... There's a real diff panel per project now: branch

... agents and forgetting manual Git commits during rapid iterations. Git is great for milestones, but it fails as a real-time safety net. So, I built LocalHub . It s a 100% local, background version control system that runs inside VS Code. Here are the core facts about the architecture: 7 Smart Triggers: It automatically snapshots files on save, tab switch, editing pause, focus loss, and external changes. No git add or git commit needed. Agent Diary (AI Protection): It logs ... ... agent actions separately. Crucially, AI cannot empty the trash or delete your history. If an agent hallucinates, you can instantly roll back. Ultimate Recovery: We built dedicated trash bins for everything deleted files, "Zombie Branches

... graveyard folder. It's embarrassing but it saved me twice. Got me thinking: why hasn't writing software figured this out? Git has solved this problem for code for like 20 years. You get a full timeline, branches, diffs line by line. But others (non-devs), still doing "copy the whole file." I started prototyping something simple, not the full complexity of Git, just the core ideas applied to writing: A scrubable timeline. Slide back to any point in your ... ... document's history, not just "undo" but days ago. Branches

... everyone, Sam here. Tomorrow I m launching nodal on Product Hunt. nodal is a macOS menu bar app that automatically tracks the time you spend working based on the Git branch you re active on. I built it because I ve always hated manual time tracking. As a developer working across multiple repositories and branches

... opening the flow editor once. Here's what happened: We have a YouTube script generation pipeline on Serenities. It takes a keyword (or YouTube URLs), generates  an outline, then writes each section with precise word counts. Two branches one for URLs, one for direct topics. The URL branch

... smart" to be useful. Comment from Navanita Devi(@navanita_devi): Dedy's point is interesting, the opacity problem is real. Most people conflate "AI having access" with "AI having unchecked control." They're very different things. Read-only access vs write access, logged actions vs

... lock test files from agent edits completely Comment from Maksim Ryndin(@maksim_ryndin): Front-end: mostly @Gemini , a bit of @Claude Code . In browser chats, prefer avoiding to give any access to the operating system. Backend: with @VS Code , usually with Rust (even for browser extensions like https://semplico.app/sleach ), mostly myself Comment from Tom Eldridge(@teldridgeldn): Running Claude Code inside Cursor and as a VS

... than what I had built over the last 3-4 weeks. Here are some takeaways I have till now (Claude Code based): A project management workflow with tickets/issues works really well, just like usual software development process Use Git branching to experiment on ideas and merge or not (if you want to throw an experiment) Use git worktrees if you want to work on multiple tasks in parallel - more advanced, I do not use this Use Claude.me to document your ... ... software workflow, code formatting, linting standards, etc. Ask Claude to follow Claude.md when tackling tasks - Claude Code can access GitHub issues, manage branches

... Cool you've built on both! What did you learn from the experiences, any major (or minor!) takeaways? Comment from David Justice(@david_justice): I love Neon! Idk why more people don't talk about it. Easily create branches and revert back to a time in history if you accidentally drop tables or needed data. Also have one of the most generous free tiers and ui is very clean. Comment from Tim Su(@iambetaful): We use Neon, but Supabase seems ...

... software engineering, trunk-based development (TBD) is a strategy to manage git workflows that favours merging commits into the `main` branch even before the feature is fully complete, instead of the git flow that favours working on long-lived branches