Fetching from the wire…
OSS2026-07-28 · source-backed
209 stars, 190 commits, posted as a Show HN July 26. Distills production agent traces into smaller open models via the Tinker API and exposes a routing endpoint dispatching between frontier and distilled models per request, claiming RouterBench results maintaining frontier quality at 27% lower cost, with optional closed-loop simulation training and a wmo optimize distill command that adds compressed models to a local pool as traces accumulate. Early, but it's the clearest packaged version of a loop plenty of people have hand-rolled. (GitHub)
Each link below shares sources, entities, or timing with this story.
I check Product Hunt maybe once a week and usually regret it. Today's board is worth reading as market structure. The July 30 leaderboard: SKI at 277 upvotes (free voice input for Claude Code and Codex). AI Search Console at 249 (prompt analytics and citation mapping). Memmy A...
A Rust/io_uring storage engine posted to Show HN today moves KV cache off the GPU into a petabyte-scale store co-located on GPU hosts, so a prefix computed on one GPU serves to others (GitHub). Claims million-plus IOPS within 1ms and 66x TTFT speedup on cache hits at 128K cont...
One Python file wiring a model to exactly one tool, bash, with the system prompt loaded from a file the agent can rewrite. Tools, memory, skills, conventions all have to be grown by the agent session by session into a self/ directory that persists. Model access goes through Si...
An Apache-2.0 OpenAI-compatible gateway at 695 stars, with per-user and per-agent spend controls, that ingests OpenTelemetry traces from existing agent workflows, builds a simulation from them, and uses that to fine-tune an open-source model you own (GitHub). Rust-native, repo...
GitHub put per-team model policy targeting into public preview on July 31, letting an org allow different model sets to different teams instead of one org-wide list. Same week, Cursor shipped admin controls over its Router restricting which of Cost/Balance/Intelligence modes t...
Three signals in ten days: Amp published "Who Cares About the Model?" on July 29 describing swapping its default model overnight with no user complaints; Cursor Router now analyzes each request and picks a model under Intelligence, Balance, or Cost objectives; and GitHub retir...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.