Ben Vinegar
@benv.ca
Co-founder modem.dev. Before: founding engineer then VPE at @sentry.io. Organizes Devtools Toronto. 🇨🇦 More: benv.ca
It's becoming normal practice to plug your coding agent into tools like Slack and helpdesk software to find customer insights. Turns out it's super inefficient, because you're just fetching and smashing raw strings into your context repeatedly. So we released a new MCP server that solves that.
✅ Yes the code matters In experiments we ran, well-named and organized code can dramatically improve agent outcomes: - fewer tokens spent/more context window preserved - fewer confidently wrong answers developed Verified against 14 model/harness combos incl Sol, Fable, GLM 5.2 and more
Ben Vinegar@benv.ca · 2w ago👋 New blog post: "How coding agents read files – and how to write for them" In it I explain how code optimized for agent-discovery can lead to major token savings on search + retrieval, even with frontier models like GPT-5.6 Sol Plus other important findings ⬇️
👋 New blog post: "How coding agents read files – and how to write for them" In it I explain how code optimized for agent-discovery can lead to major token savings on search + retrieval, even with frontier models like GPT-5.6 Sol Plus other important findings ⬇️
🎙️ State of Agentic Coding #8 with guest Mario Zechner is out. In this episode: - Vibe-checking Fable & GLM 5.2 - @mitsuhiko.at teaches us RL - How harnesses smooth out model jank - Loops, surviving AI fomo, and more YT: www.youtube.com/watch?v=_lfp... Spotify: open.spotify.com/episode/6m40...
🛠️ hunk 0.17.0 is out Patches from 10 new contributors, w/ small fixes & quality-of-life improvements like: - official homebrew install (brew install hunk) - launch hunk w/o menu bar - better session reliability - esc no longer quits, proper transparent bg colours, & more
Just noticed a vibe-coded project of mine cost $200/month to serve a few dozen users/day 😭 Turns out it was doing 10s of millions of unnecessary Redis polls/day, even when idle - telemetry alone cost $$ a possible clue for why infra providers are growing so much right now
🎨 hunk v0.15.0 is out: - supports git --color-moved - transparent backgrounds - color token fixes - more themes - Sapling support (it's a VCS from Meta) also now there's a website → hunk.dev
A pattern I'm seeing with AI debugging: it's easy to get stuck inside the model's search space. So you burn tokens & time chasing candidate fixes, while the real answer sits in context only you have ... but never explored, because you quietly surrendered your thinking.
hunk 0.14.0 is out w/ lots of user contribs + fixes: 🆕 expand/collapse inline source (vid 👇) 🆕 custom themes 🆕 new theme: catpuccin latte/mocha 🆕 select + copy to clipboard 🆕 renders wide chars like 你好 & emoji properly
As expected, another tool by @benv.ca has become part of my workflow - Glance solves the problem of given terminal based agents access to screenshots. It's been super helpful while I'm building by first TUI glance.sh
glance.sh — Share an image with your coding agent
Paste a screenshot, get a temporary URL. Built for terminal environments, coding agents, and CLI workflows where copy/pasting images is hard.
glance.sh
In 3 weeks at the next Devtools Toronto: traces.com - share agent coding sessions polarity.so - sandboxes for agent testing superuser.app - agents in your team chat tidra.ai - AI code maintenance at scale docsalot.dev - docs agents can read codepathfinder.dev - trace vulns across your codebase
State of Agentic Coding #5 with @mitsuhiko.at is live. In this episode: - death of the IDE premature? - what's with all the bugs - will future models fix yesterday's AI code? - token abuse, slowing down, and more! YT: www.youtube.com/watch?v=qCpY... Spotify: open.spotify.com/episode/7wUV...
My terminal diff tool, hunk, has added a lot of good stuff: - horizontal scrolling w/ mouse support - pinned file headers - way better file browser UI - draggable scrollbar - perf & bug fixes ⬇️ Here's what it looks like today
Ben Vinegar@benv.ca · 5mo agoI've been building a terminal diff tool for agentic code review: - agent comments + agent remote control - syntax highlighting - auto-responsive layouts (stacked & split) - file browser - mouse support - uses OpenTUI + Diffs from Pierre Computer Co
Montreal's Metro is the 4th-busiest rapid transit system in North America But for brutalist vibes, it's unmatched
I've been building a terminal diff tool for agentic code review: - agent comments + agent remote control - syntax highlighting - auto-responsive layouts (stacked & split) - file browser - mouse support - uses OpenTUI + Diffs from Pierre Computer Co
Coding was never the slow part. That's why we built Modem, your dev team's auto-triage PM. It listens to user signals 24/7, finds problems before you do & does the in-between work – filing tickets, pinging teammates, closing the loop. And it's not vaporware! You can try it out today at modem.dev
How I vibe-eval new model releases: - Am I spending more time building because it's more fun? - Can I give higher-level guidance than before and it works out? Only a few days in, Opus 4.6 clears both. 🤷
🇨🇦 Devtools Toronto #6 – the version where we just hang out – returns Tuesday, Feb 17th at Mascot Brewing 👇 RSVP below
🇨🇦 Devtools Toronto #6 – the version where we just hang out – returns Tuesday, Feb 17th at Mascot Brewing 👇 RSVP below
At Modem we've gone all-in on database branching to accelerate our development – sanity testing migrations, making sure preview builds work end-to-end, and more. 👇 The folks at Neon thought we were pushing the envelope enough to merit writing about it. Worth a read! neon.com/blog/behind-...
Behind Modem.dev’s Product Velocity - Neon
Modem is building an auto-triage, PM agent that centralizes scattered user feedback and automates follow-ups. Powered by Neon branching.
neon.com
Weekend thoughts on Gas Town, Beads, slop AI browsers, and AI-generated PRs flooding overwhelmed maintainers. I don't think we're ready for our new powers we're wielding. lucumr.pocoo.org/2026/1/18/ag...
Agent Psychosis: Are We Going Insane?
What’s going on with the AI builder community right now?
lucumr.pocoo.org
It was a good year: 1. Started a company (modem.dev), raised some money, hired a great team 2. Hit 1.9k GitHub stars on Counterscale (w/ @stordahl.dev) 2. Hosted DevtoolsTO 3x 4. Invested in some great 🇨🇦 startups 5. Visited Japan & Austria 6. Took some pics 📸
Blue Prince was my GOTY and I think nerdy programmer types here would enjoy it. Also it plays well on Linux via Steam's Proton compatibility layer.
Blue Prince@blueprincegame.bsky.social · 9mo agoWe are thrilled to have been awarded Best Indie Game at the Golden Joystick Awards. 🏆 Congratulations to all the winners and to everyone nominated. 2025 has been an incredible year for games and we are delighted to have been included alongside so many amazing titles.
