🛠️ Hacker News Tools

Show HN: Relaymux, a tmux-based meta-harness for local coding agents Hey HN,<p>There’s been a lot of interest recently in meta-harnesses, loops, and multi-agent orchestration. Obviously, there are already a lot of good tools: Conductor, cmux, the native Codex &#x2F; Claude Code apps, etc.<p>For my own use cases, I’ve felt that the orchestration layer tends to feel overengineered. I mostly wanted a simple local harness (i.e Pi) for running and tracking CLI agents with the ability to hop in (via tmux). Relaymux is my opinionated attempt at that.<p>A few design principles:<p>- The frontend is just Telegram &#x2F; iMessage &#x2F; CLI. If I want more visibility, I hop into tmux.<p>- Subagents are normal interactive CLI agents running in tmux windows, usually with their own worktrees.<p>- The harness owns the tmux session, so each longer task becomes a named tab&#x2F;window. Subagents report back to the orchestrator via CLI when they’re blocked or done. Then the orchestrator just messages me on Telegram &#x2F; iMessage<p>- It works with any CLI agent that has an interactive terminal mode, so I don’t need special print-mode&#x2F;non-interactive support. This means I don’t need to stress about the Agent SDK &#x2F; claude -p billing limitations.

Agentic Resource Discovery Specification <a href="https:&#x2F;&#x2F;developers.googleblog.com&#x2F;en&#x2F;announcing-the-agentic-resource-discovery-specification&#x2F;" rel="nofollow">https:&#x2F;&#x2F;developers.googleblog.com&#x2F;en&#x2F;announcing-the-agentic-...</a>

Show HN: An 8-bit live gamecast for baseball Hey HN, I built a website to watch live baseball games in an 8-bit broadcast. It takes live MLB data streams and converts them into near real-time pixel art gamecasts.<p>Been waiting to share this for when there’s actually a good slate of games happening since the site is pretty bare otherwise.<p>Here is today&#x27;s schedule:<p>Mets @ Reds - 9:40am PDT <a href="https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;824503" rel="nofollow">https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;824503</a><p>Royals @ Nationals - 10:05am PDT <a href="https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;822721" rel="nofollow">https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;822721</a><p>Marlins @ Phillies - 10:05am PDT <a href="https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;823450" rel="nofollow">https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;823450</a><p>Tigers @ Astros - 11:10am PDT <a href="https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;824178" rel="nofollow">https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;824178</a><p>Padres @ Cardinals - 11:15am PDT <a href="https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;823044" rel="nofollow">https:&#x2F;&#x2F;ribbie.tv&#x2F;watch&#x2F;game&#x2F;823044</a><p>..and another 14 games throughout the later day.<p>I&#x27;m still early on in this project, but I&#x27;ve tried to add little details with actual stadiums, day and night modes, between inning graphics and interstitials, live scoreboards, etc.<p>Would love any feedback and ideas. Thanks for checking it out!

Show HN: Vpod – Tiny Linux sandbox running in WASM Hi HN,<p>I spent the last few months reading the RISC‑V specification to build the lightest possible sandboxes. The idea behind a vpod is to quickly spin up a Linux sandbox from snapshots (Alpine by default) without any setup or subsystem required.<p>The trade-off for portability and security is raw CPU speed. So we don&#x27;t expect it to match native workloads with Python or pip, for example.<p>More info is in the README <a href="https:&#x2F;&#x2F;github.com&#x2F;capsulerun&#x2F;vpod" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;capsulerun&#x2F;vpod</a><p>Happy to answer any questions!

Show HN: Zkit – Go libraries for building agents, not a framework

Show HN: In-browser Python/Pandas/Git practice with animated Git simulator I&#x27;ve created an in-browser Python&#x2F;Pandas&#x2F;Git practice environment for my online learning platform and also for my corporate training classes. I&#x27;d be happy to discuss how I went about designing this, how I&#x27;m using it in my classes, and the architectural decisions I&#x27;ve made.<p>Most interesting, to me, is how much is running in the browser. Thanks to Svelte, Pyodide, isomorphic-git, LightningFS, and CodeMirror I&#x27;m able to provide a full environment for Python, Pandas, and Git.<p>I built much of this with Claude Code, and I&#x27;m happy to discuss how that went — what worked well and where I had to step in and make the calls myself.<p>I&#x27;m especially excited about the Git simulator: it shows the commit tree change as you run commands, plus an animated view of how files move between the working tree, the staging area, and HEAD.<p>The AI tutor, which uses Claude Haiku, was given my newsletters, classes, and exercises as inputs, along with a description of my pedagogical approach: instructors should give hints and feedback, but not reveal the answer.

TREX: An AI code reviewer that runs your code

Agentic coding deserves more than a chat box bolted onto VS Code

French physicist and media star loses doctorate after plagiarism investigation

Occupancy Math on the AMD MI355X: A From-First-Principles Guide

AI demands more engineering discipline. Not less

ICE Appears to Be Buying Immigrants' Tax Identifiers from a Data Broker

Computed goto for efficient dispatch tables (2012)

Show HN: Mira – Open-source and self-hosted AI code reviewer Hey HN,<p>I&#x27;m Jay, co-creator of Mira. An open-source, self-hosted AI code reviewer where you BYOK (bring your own key).<p>Local models are getting really good.<p>Hosted frontier models are getting really expensive.<p>So, we built Mira. Mira has some snazzy features too:<p>- It&#x27;s really fast at reviewing. Average is 77s compared with Greptile&#x27;s 5 minutes. Your PRs aren&#x27;t going into a queue on a cloud somewhere.<p>- Mira performs a blast radius and see what damage will be done with code change.<p>- Learns your codebase&#x27;s patterns from the repo itself and enforces them without a config file.<p>- It&#x27;s self-hosted, so your code never leaves your infrastructure. Bring your own model - OpenAI, Anthropic, or a local LLM behind your firewall.<p>We&#x27;re really active, and looking for active contributors. Come join us in our discord!<p><a href="https:&#x2F;&#x2F;github.com&#x2F;miracodeai&#x2F;mira" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;miracodeai&#x2F;mira</a><p>Thanks, Jay

Show HN: Ferrix AI – Agentic Product Management Platform Hi HN, for the past few months, we’ve been working on Ferrix AI (<a href="https:&#x2F;&#x2F;ferrix.ai&#x2F;" rel="nofollow">https:&#x2F;&#x2F;ferrix.ai&#x2F;</a>)<p>As AI agents speed up engineering, deciding what to build has become the bottleneck. Developers got faster because agents fit into their workflow: tech design, code review, testing. They still make the key decisions, while AI handles mechanical work around them.<p>At Ferrix, we are automating workflow for product managers to make product decisions efficiently. Agents handle the research synthesis, spec writing, and progress tracking, while PMs, designers, and engineers collaborate in a shared workflow.<p>Ferrix AI is live, and you can start free → <a href="https:&#x2F;&#x2F;app.ferrix.ai&#x2F;" rel="nofollow">https:&#x2F;&#x2F;app.ferrix.ai&#x2F;</a>

Human Judgment as a Specification

The discovery that changed how scientists think about memory

RFC 10008: The new HTTP Query Method

U.S. Science Is in Chaos

GLM-5.2 is the new leading open weights model on Artificial Analysis