Fetching from the wire…
Public story · 2026-09-09 · high
It hands verification to KiCad's own electrical and design-rule checks rather than an LLM judge, and reports its GitHub traction plainly.
Why now: Copperhead posted to Show HN in the September 9 coverage window with 234 points.
Copperhead runs an agent over KiCad projects by editing the s-expression source files directly rather than regenerating them from scratch. Regenerate the whole file and the diff becomes unreadable, so a reviewer can't tell what changed in a schematic or board file. Editing the source in place keeps that diff reviewable, which matters for anyone who has to sign off on a board before it goes to fab.
The tool outputs gerbers, firmware and documentation, and keeps schematics, docs and budgets in sync as it works. For verification, it doesn't ask an LLM to judge whether the board is correct. It hands that job to KiCad's own electrical rule check and design rule check.
Traction is small and reported without inflation: 195 GitHub stars, 513 monthly npm installs, 234 points on Show HN. The CLI is Apache-2.0 licensed. Cloud pricing is $49 per user per month, free for open hardware projects.
A lot of agent tooling reaches for an LLM to check its own work. Copperhead reaches for the domain's existing deterministic checker instead, so a passing run means something a human already trusts before the agent ever touched the file. That's the part worth copying for anyone building an agent over a structured source format. Ask whether the tool edits in place and verifies with the format's own checker, or regenerates and hopes an LLM catches what it broke.
Each link below shares sources, entities, or timing with this story.
Allen Bargi's August 15 post hit 302 points arguing that AI collaboration rewards context-sharing, examples, and feedback over precise instruction (Hacker News). The pushback holds that the piece conflates management with leadership. mikeocool calls it "the most low effort ver...
Simon Willison highlighted Sophie Alpert's short internal policy on AI-assisted writing, whose load-bearing rule is accountability: "You must stand behind every idea and every sentence in your docs," because you can't excuse a bad passage by blaming the LLM, and doing so waste...
Simon Willison released it August 4, calling it "the most significant new version since the initial launch of the project," which from him is not marketing. The agent-relevant pieces: tools can raise llm.PauseChain to stop for human approval, and chains resume from pending cal...
Simon Willison's release scans a JSON structure for duplicate strings and substrings from a supplied replacements object, swapping matches for compact references, marking condensed regions with {"$r": [...]} and pointing at replacements via {"$": "ID"}. Fully reversible via un...
Simon Willison shipped a PauseChain exception to cleanly pause a tool chain for human approval, guaranteed unique tool_call_ids (synthesizing ULIDs when providers omit them), and resume-from-history support. He says Fable produced the API design, tests, and docs across both LL...
0.35 adds gpt-6-astra to the CLI's OpenAI provider, so llm -m gpt-6-astra works against the same logging, template and fragment machinery as every other model in the tool. For anyone scripting cross-model evals, that means a new frontier model needs zero new plumbing to enter...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.