Chunk sidecars
Hi HN! My name is Olaf, I work at CircleCI as a technology advisor in the CTO office, came in through the acquisition of my company Vamp.io (progressive delivery for microservices on k8s) in 2021. Wanted to hear the HN community feedback and thoughts on a project we think could be very interesting when adding AI coding agents to the SDLC and your CI pipelines. Our team at CircleCI built Chunk sidecars after repeatedly running into the same issue internally: by the time our CI catches a failure, the agent has already moved on and most of the useful context is gone. The basic idea of Chunk…
Chunk sidecars launched on May 27, 2026 with 77 votes, #492 of 20,538 launches that month and more than 95% of that year's launches.
What it does
AI agents write code fast. Validation still happens after the push — by then the context is gone. Chunk sidecars run scoped microbuilds before commit, in a real CI mirror. Auto-detects your stack. ~27s average vs ~5 min billable compute for a full run. 3x–5x fewer tokens in retry loops. If something fails, the agent iterates before anything reaches shared CI. Run chunk init. Works with Claude Code, Codex, Cursor, or custom agents. Free for all CircleCI users.
Chunk sidecars bring lightweight validation into the inner loop, so AI coding agents catch failures locally before they ever reach CI.
Share on X Share on LinkedIn Share on Facebook Share on Reddit Share on Hacker News 🎉 Update 5/22: Chunk sidecars are now available to all CircleCI users, including Free plans. Local development and remote validation were always meant to work together: developers iterate on their machine, run a few manual checks, then push to CI to clear code for production. But AI development broke that balance, flooding CI with a volume of commits no developer has read, let alone tested. Chunk sidecars restore the balance: lightweight, preconfigured environments that run alongside your local workflow and validate changes as they happen. Agents get actionable feedback while they work, and CI stays focused…from circleci.com
In the maker’s words, at launch
Hi HN! My name is Olaf, I work at CircleCI as a technology advisor in the CTO office, came in through the acquisition of my company Vamp.io (progressive delivery for microservices on k8s) in 2021. Wanted to hear the HN community feedback and thoughts on a project we think could be very interesting when adding AI…
Does a similar job
all alternatives →
Huzzah – a novel approach to coding with AIAug 2026 · danielvaughn.dev · ▲384Hello everyone. I've been working on this experimental editor called Huzzah. I've been working almost exclusively with coding agents since January of this year, and over the past few months I began to feel utterly exhausted by them. They're great, but I'm finding it more and more tedious to write full sentences for every change I want. Not only that, but it seems there's a complexity limit for codebases - beyond a certain point the agent begins confusing itself. I'd like to go back to writing code, but I don't want to go all the way back to fully manual coding. So I've come up with this…
BulletAug 2026 · codewithbullet.com · ▲240Hi HN, excited to be sharing this with you guys today. TL;DR: Bullet is a fast coding agent. At our company, we were burning hours waiting on agent runs. The models weren’t the problem. The agent loops around them were slow (unnecessary planning, oversized context, repeated searches, and serial tool calls). So we built Bullet as a passion project, the goal being to help you iterate faster, without giving up quality: - It automatically selects the model and reasoning level for each prompt. - It runs independent searches, reads, and commands in parallel. - It uses targeted code search instead…
ReplicasJun 2026 · replicas.dev · ▲239Hey HN! I'm Connor, creator of Replicas. I'm a heavy Codex/Claude Code user, and I wanted to hand off tasks to agents like I would to an engineer - kick something off before work, let it scaffold while I'm at lunch, even come back to a ready PR. The problem: existing alternatives use workspace VMs that are black boxes. I have no access to any of them - they can't run tests, read logs, or use tools like the Playwright MCP (which I use to verify agents actually fixed what they claim to fix). Replicas gives agents real cloud workspaces where they can: - Run your test suite and read the…
Parallel CodeMar 2026 · parallelcode.app · ▲159I kept running into a structural problem with AI coding tools. You try one agent. It makes changes. You try another. Now you're manually diffing branches and cleaning up experimental commits. The tools are powerful, but the workflow isn’t designed for comparison. So I built Parallel Code. It runs Claude Code, Codex, and Gemini CLI natively (no API wrapper, no feature abstraction), but isolates each agent in: its own terminal its own Git worktree its own feature branch The workflow becomes: Spawn N worktrees Assign one AI agent per worktree Let them implement independently Compare diffs Merge…
ViberiaMay 2026 · getviberia.com · ▲146Hey HN, This is my take on the agent harness. Everything on an isometric map. Agents are grouped into "buildings" that run in a sequence or a loop; e.g., the CodeForge has an agent that writes a PRD, another one that implements, and a third that reviews. Everything is customizable, you build your own buildings/teams however you want. It's a Tauri app, really light (about 8x less energy than the closest competitor I benchmarked, so it actually runs from a coffee shop on battery). It's macOS only for now, but ping me if you are willing to test the Windows or Linux version. I've been…
20+ Claude Code agents coordinating on real work (open source)Feb 2026 · github.com · ▲53Single-agent LLMs suck at long-running complex tasks. We’ve open-sourced a multi-agent orchestrator that we’ve been using to handle long-running LLM tasks. We found that single LLM agents tend to stall, loop, or generate non-compiling code, so we built a harness for agents to coordinate over shared context while work is in progress. How it works: 1. Orchestrator agent that manages task decomposition 2. Sub-agents for parallel work 3. Subscriptions to task state and progress 4. Real-time sharing of intermediate discoveries between agents We tested this on a Putnam-level math problem, but the…
More ai this month
the category →
Paintings by AI models: programs run through a simulation of oil paint.
AI · 20d ago · stillwet.art
PixVerse R2▲381A real-time world model you can explore and change
AI · 28d ago · world.pixverse.video
Source: Product Hunt launch ↗ · launch text quoted in part
Launched alongside, May 2026
the whole month →
Parallel agents, diff reviewer, and multi-model comparisons
Dev tools · May 2026 · kilo.ai
Hey HN, Henry here from Cactus. We open-sourced Needle, a 26M parameter function-calling (tool use) model. It runs at 6000 tok/s prefill and 1200 tok/s decode on consumer devices. We were always frustrated by the little effort made towards building agentic models that run on budget phones, so we conducted investigations that led to an observation: agentic experiences are built upon tool calling, and massive models are overkill for it. Tool calling is fundamentally retrieval-and-assembly (match query to tool name, extract argument values, emit JSON), not reasoning. Cross-attention…
Life & fun · May 2026 · github.com
🌱 Private, quiet space for thinking. Simple app for .md files. - zakirullin/files.md
Dev tools · May 2026 · github.com