... survey about how developers feel. Actual measurement of what came out the other end. The headline finding is the one everyone quoted: throughput is up. Epics completed per developer up 66%. Task throughput per developer up 33.7%. PR merge rate up 16.2%. Real acceleration, and I don't want to be cynical about it, because it's the thing that let me ship a product I couldn't have built otherwise. But the numbers underneath are the ones ... ... tape to the wall. Median time in review: up 441.5%. Median time to first review: up 156.6%. And this one, which is the whole article in a single stat pull requests merged

... ZonePlan SEO experiment update: merged 6 near-identical planner pages into 1 comprehensive page. Before: 4670 impressions, 5 clicks, 0.1% CTR. One day after consolidationBuilt ZonePlan (zoneplan.net), a free worldwide meeting planner. Had 6 near-identical planner pages for months. 4670 monthly impressions. Only 5 clicks. 0.1% CTR. meeting planner worldwide alone: 268 impressions, zero clicks. The fix on Aug 4: merged

... kept needing simple things like compressing an image or merging two PDFs, and almost every site asked me to upload my file to their server first. That always felt wrong to me, especially for private documents. So I built ToolHub, a small site with eight tools that run entirely in your browser. Your image gets compressed on your own device, your PDF gets merged

... editing outcomes, not architecture, and it costs me a coffee's worth of attention. Get the blueprint right and you've done 90% of your job for the week. Everything after is yeses. The agent opens PRs. It never merges. Here's the fence that makes it safe to walk away. The agent's  only  forward move is opening a pull request. It can't merge

... understanding of the codebase less than I used to - I'll touch one file and have no real sense of what else depends on it until CI (or prod) tells me. Ended up manually tracing imports before merging anything non-trivial, just to make sure I'm not breaking something three files away. Anyone else running into this, especially with AI-assisted code? Curious if it's a real pattern or just me being paranoid. Comment from Ece(@ecekc ... ... pieces connect. Comment from Alper Tayfur(@alpertayfurr): I don t think you re being paranoid. When AI writes most of the feature, we skip some of the slow work that normally builds our mental map of the codebase. Before merging

... cost per outcome. A tool that uses fewer tokens but requires multiple rounds of debugging may be more expensive in terms of developer time. I d love to see benchmarks that measure autonomy, correction cycles, and time-to-merge rather than just token consumption. Comment from Asad M.(@asadmalik901): The methodology you are looking for exists, it is just annoying to run: fix the task, vary the tool, and count total tokens until the artifact passes the same acceptance test ... ... between them as the real quality signal. A tool with a tight distribution and a slightly worse median beats one with a great median and a fat tail every time on real work. Stephanie's time-to-merge

... wrote it, so you understood it, mostly by necessity. Now understanding is a separate step you have to choose to take, not something that comes free with writing the code yourself. One habit that closes that gap fast: before merging anything non-trivial, ask the agent to explain its own implementation back to you, line by line, in plain language. Not as a formality, actually read the explanation. Two things tend to happen. Either it confirms your understanding and takes ... ... anything touching auth, state management, or async logic, the categories where "it worked in my test" and "it's actually correct" diverge the most. Curious what habits others have built for this. Do you have a personal checklist before merging

... everyone! I recently built DocuPDF, a free online PDF toolkit designed to make everyday document tasks faster and easier. You can use it to merge, split, compress, convert, rotate, organize, protect, repair, OCR, translate, and summarize PDF documents all from your browser. Some of the tools I'm currently working on include: - PDF Merge

... FetchSandbox . today i'd rather hear about you. first, thank you. for the upvotes, but honestly more for the questions/comments n the conversations about idempotency, out-of-order webhooks, deterministic replay, and what it actually takes to gate a merge were sharper than anything i'd written for myself. a bunch of you basically handed me my roadmap. I built this because.. i kept watching ai agents write stripe and auth integrations that passed every test and then broke ... ... receipt showing it either survived or exactly where it broke. no keys, no account, no staging wait. that's the whole loop. if you're on a team shipping real payment or auth integrations and want proof before merge

... started asking two things instead: how confident am I it works, and how bad is it if I'm wrong. That splits everything into three lanes: Zero blast radius (docs, copy, an endpoint nothing depends on yet): the agent merges it. If it's wrong, I revert it - costs me nothing. Most feature work: a QA agent buys the confidence instead of me. It opens the preview in a real browser and clicks through the change like an impatient user ... ... browser, and files a ticket automatically so a transient blip fixes itself and I'm not the smoke detector. Honest caveat: this is safe because I'm solo with a small user base and instant rollback. If a bad merge

... worker. For UI testing, a remote headed browser is not just a convenience. It changes the debugging loop. Instead of asking "what did the script think happened?", you can ask "what did the user actually see?" ### Harness Code vs Local Agent Harness Another important decision is where the agent capability should come from. One option is to build the entire reasoning layer yourself: call an LLM API directly, manage prompts, retries, tool calls, screenshots, JSON parsing, and safety rules. That ... ... gives maximum control, but it also means the test harness becomes an AI platform project. The alternative

... genuinely relevant, researching what is happening inside them, finding the right person, verifying that the contact information is real, understanding why they might care right now, writing something that does not look like it came from a mail-merge factory. Yet most AI SDRs focus on sending more emails with fewer humans involved. We took the opposite approach with Causo. The system does the annoying work: searching live sources, cross-checking companies, researching why they fit, finding decision-makers, verifying ...

... device since it's all on-device inference. The interesting technical challenge was the clustering logic. I ended up using a union-find (disjoint set) structure combined with NaturalLanguage embeddings to figure out which notes are actually related, then merge them into clusters that update as you add more notes. Getting the similarity threshold right so Clouds don't over-merge

... rate and the drop opened an incident which, in turn, dispatched Claude Code in the cloud, and a few minutes later a pull request appeared that restored the cache structure the refactor had broken. A developer reviewed the change, merged it, and the hit rate recovered within the hour. Without that loop, a regression like this tends to pop up weeks later in a cost report, and someone spends an afternoon dissecting deploys to find it. Latitude was built around ... ... traffic behind the Signal becomes a dataset, and a regression test in your repository, not in Latitude, replays that dataset against the agent as an ordinary CI check on every pull request. The fix has to pass it to merge

... building a multi-agent ecosystem in the background. The Unbreakable Pipeline: A PM types a spec in Solidea An agent translates it into a Draft Architecture in Solarch The CTO approves it Code is deterministically generated and merged Solidea s Marketing Agent detects the merge

... Dependency Watchdog : A daily security scanner for your GitHub repos that reads dependency files, cross-references versions against CVE databases and GitHub Security Advisories, and delivers a prioritized Telegram digest grouped by severity - with memory to suppress repeat alerts. Merge Conflict Resolver : Provide a consistent workflow for resolving merge

... Cursor just dropped " Origin ," a GitHub competitor. Cursor is moving way beyond "VS

... Zaitzow(@dzaitzow): @anna_ludwinowski I guess if I really think about it - it helps solve the 1 to many problem for lead capture - making every lead get something really unique at the end of the intake survey / quiz funnel VS being bucketed in a segment that MAY apply to them and their specific circumstance. Here is the product and we're live today if anyone feels like poking around! @stacywycof83995 @oshylabs @thenameisarian @virko_kask if you feel like checking ... ... Comment from Anna Ludwinowski(@anna_ludwinowski): @getosmo  "The receipt IS the scope" - YES, that's a sharper way to collapse messaging and product into the same thing. Most people treat those as two separate problems (what we say vs

... Something I keep running into while building LineageLens: teams know they should review AI-generated code before it merges, but "reviewed" in practice means "someone clicked Approve." The diff could have been open for 3 seconds or 30 minutes. The record looks the same. I shipped a feature this week that tries to make this distinction real. Three behavioral signals time-per-line on the diff, comment count, lines reviewed get scored into a classification: `shallow`, `adequate`, or `deep`. Anything ... ... below 1 second per line gets hard-classified as `shallow` regardless of score, because no one is actually reading at that speed. The classification gets signed with Ed25519 and stored as an attestation. A CI gate endpoint then blocks merges

... live on PH Aug 22, so this thread's been good practice at saying it in one paragraph.) Comment from Demetrio Koro(@aktivist): siOsi upload one selfie and it reads your 12-season color type, undertone, gold vs silver, and which hair colors suit you, then grades the actual makeup in the photo (flash cast, oxidation, coverage), which most tools skip. Free, no signup. Built it because every beauty sub has a queue of "what season am I" posts waiting ... ... Gemini, since more buying decisions now start there instead of traditional search. If you want to see how an AI agent currently reads your own site, we put out a free tool that shows what an AI fetches vs

... time. What I'm genuinely curious about: what does your team produce today as evidence when AI code ships? Is there anything machine-verifiable in that process? And where do you think the enforcement gate should actually live pre-merge block, or post-merge

... during the day, each with enough context that an agent can finish one without coming back to me for a product decision. The ready ones get a label. From there, the agents carry each issue through implementation, review, fixes, merge, and deploy checks, then unblock the next one. I stay in charge of what gets built. They do the rest. A few things that made it work, not just demo: The decision of what to build stays human. Only ... ... pointed at newsletters instead of pull requests: say what you want once, and let the system handle the repetitive parts. If you're building with agents, I'd genuinely like to hear where you keep yourself in the loop versus

... workflow and always feel like I'm missing out. What kept happening to me: 3-5 agents fan out, each opens a PR CIs pass, diffs look reasonable I can't possibly click through every preview by morning I merge based on the diff and CI signal Something breaks in prod the next day because the agent did what I asked literally, but the feature doesn't REALLY work CI passing doesn't tell you whether clicking the button does ... ... anything. The more agents I run in parallel, the more PRs I'm merging

... analyze, tests and security scans on every PR so I can't merge something dumb at midnight Golden tests (Flutter's screenshot testing) - half tool, half safety net, but I couldn't touch a chart without them also +1 on gmail as a reminder system, my drafts folder is basically a todo app at this point Comment from Richard Acheampong (@richatsealedvault): Sealit password manager, Claude code, VS

... Something I built led to a design decision I want to get feedback on. LineageLens is a free VS Code extension that captures every AI code insertion and scores it for risk on a 0 100 scale. Works with Cursor, Copilot, ClaudeCode, Gemini CLI. Zero config on install just start using your AI tools and your insertions start showing up in the sidebar. The scoring is deterministic rules: +28 for credential-like material, +24 for eval/exec patterns, +22 for subprocess ... ... intentionally mixes both because incomplete capture weakens auditability, but I agree there s a conceptual difference between this code appears dangerous and we cannot fully explain how this code was generated. Exposing those as parallel dimensions instead of one merged