Robin/openwiki-wiki
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore: daily update 2026-08-14

Browse Source
This commit is contained in:
OpenWiki Bot
2026-08-14 02:18:06 +08:00
parent 1737454ad1
commit 021e929b92
26 changed files with 290 additions and 154 deletions
Download Patch File Download Diff File Expand all files Collapse all files
sources/hackernews.md
+44 -4
View File
@@ -2,7 +2,7 @@
type: "Reference"
title: "Hacker News — evidence index"
description: "Compact evidence trail for the Hacker News connector (hackernews-1, Robin-curated life-sciences + AI workflow). Synthesis lives on /themes.md, /commitments.md, etc.; this page only records ingestion coverage and notable items."
tags: [hackernews, evidence-index, source, 2026-08-12, 2026-08-10]
tags: [hackernews, evidence-index, source, 2026-08-13, 2026-08-12, 2026-08-10]
---
# Hacker News — evidence index
@@ -15,10 +15,50 @@ Compact evidence trail for the Hacker News connector (`hackernews-1`, Robin-cura
- **Backend:** public HN feeds (`top`, `new`) + Algolia HN search queries
- **Auth:** none — public feeds
- **Window:** last 24 hours
- **Latest fetch:** `2026-08-12T18:00:30Z` (LLM-agent topic set: `organoid`, `CelVivo`, `3Brain`, `Cellink`, `openwiki`, `deepagents` — 6 queries; the source-specific instruction "Track organoid / CelVivo / 3Brain / Cellink / openwiki / AI agent stories" is the active query set; prior 2026-08-10 18:00, 2026-08-08 18:00, 2026-08-08 03:15, 2026-08-07 12:29, 2026-08-03 18:00, 2026-08-02 18:00, 2026-08-01 18:00, 2026-07-31 18:00, 2026-07-30 18:00, 2026-07-29 18:00, 2026-07-28 18:00, 2026-07-27 18:00, 2026-07-26 18:16 and 18:00 pulls also documented below)
- **Raw dumps:** `~/.openwiki/connectors/hackernews/raw/2026-08-12T18-00-01-956Z/hackernews-results.json` (current — one new low-engagement organoid signal, see 2026-08-12 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-10T18-00-02-327Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-10 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-08T18-00-01-989Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-08 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-08T03-14-40-856Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-08 03:15 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-07T12-29-44-173Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-07 entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-03T18-00-02-766Z/hackernews-results.json` (earlier — material findings, see 2026-08-03 entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-02T18-00-02-278Z/hackernews-results.json` (empty) · `~/.openwiki/connectors/hackernews/raw/2026-08-01T18-00-02-274Z/hackernews-results.json` (empty) · `~/.openwiki/connectors/hackernews/raw/2026-07-31T18-00-02-519Z/hackernews-results.json` (empty) · `~/.openwiki/connectors/hackernews/raw/2026-07-30T18-00-02-784Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-29T18-00-02-667Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-28T18-00-02-178Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-27T18-00-02-615Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-26T18-16-06-805Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-26T18-00-02-441Z/hackernews-results.json`
- **Latest fetch:** `2026-08-13T18:00:30Z` (LLM-agent topic set: `organoid`, `CelVivo`, `3Brain`, `Cellink`, `openwiki`, `deepagents` — 6 queries; the source-specific instruction "Track organoid / CelVivo / 3Brain / Cellink / openwiki / AI agent stories" is the active query set; prior 2026-08-12 18:00, 2026-08-10 18:00, 2026-08-08 18:00, 2026-08-08 03:15, 2026-08-07 12:29, 2026-08-03 18:00, 2026-08-02 18:00, 2026-08-01 18:00, 2026-07-31 18:00, 2026-07-30 18:00, 2026-07-29 18:00, 2026-07-28 18:00, 2026-07-27 18:00, 2026-07-26 18:16 and 18:00 pulls also documented below)
- **Raw dumps:** `~/.openwiki/connectors/hackernews/raw/2026-08-13T18-00-02-662Z/hackernews-results.json` (current — one directly-relevant `agent-memory-architecture` evidence (MCP Memory using OKF), see 2026-08-13 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-12T18-00-01-956Z/hackernews-results.json` (prior — one new low-engagement organoid signal, see 2026-08-12 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-10T18-00-02-327Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-10 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-08T18-00-01-989Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-08 18:00 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-08T03-14-40-856Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-08 03:15 UTC entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-07T12-29-44-173Z/hackernews-results.json` (prior — no promotable hits, see 2026-08-07 entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-03T18-00-02-766Z/hackernews-results.json` (earlier — material findings, see 2026-08-03 entry below) · `~/.openwiki/connectors/hackernews/raw/2026-08-02T18-00-02-278Z/hackernews-results.json` (empty) · `~/.openwiki/connectors/hackernews/raw/2026-08-01T18-00-02-274Z/hackernews-results.json` (empty) · `~/.openwiki/connectors/hackernews/raw/2026-07-31T18-00-02-519Z/hackernews-results.json` (empty) · `~/.openwiki/connectors/hackernews/raw/2026-07-30T18-00-02-784Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-29T18-00-02-667Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-28T18-00-02-178Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-27T18-00-02-615Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-26T18-16-06-805Z/hackernews-results.json` · `~/.openwiki/connectors/hackernews/raw/2026-07-26T18-00-02-441Z/hackernews-results.json`
## Latest update — 2026-08-12 18:00 UTC
## Latest update — 2026-08-13 18:00 UTC
- **Coverage:** 2 feeds (`top`, `best`), 6 search queries (the same simplified LLM-agent topic set from the source-specific instruction "Track organoid / CelVivo / 3Brain / Cellink / openwiki / AI agent stories": `organoid`, `CelVivo`, `3Brain`, `Cellink`, `openwiki`, `deepagents`). `top` feed returned 25 items, `best` feed returned 7; deduplication gave **27 unique items across both feeds** (5 overlap: `49285244` DeepSeek Harness developer preview, `49281916` Codex in ChatGPT desktop app for Linux preview, `49284774` Deutsche Bank yuan clearing, `49286341` Spaghettifying DRAM, `49285982` Gloomberb). The 6-query lane produced the hit table below; the connector completed normally at `fetchedAt: 2026-08-13T18:00:30.603Z` (24-hour window).
- **Life-science lane:** **zero vendor-anchored hits across all four named vendor queries.** All six Algolia searches (`organoid`, `CelVivo`, `3Brain`, `Cellink`, `openwiki`, `deepagents`) returned `nbHits=0` in the 24 h window. A title/URL regex sweep across the 27 unique top+best items also turned up zero organoid / 3D-bioprinting / CelVivo / 3Brain / Cellink / ClinoStar / ClinoReactor / MEA / spheroid / organ-on-chip items. The "no material HN footprint for the 4 vendor lines" finding continues to hold — this is the third pull since 2026-08-07 to have a fully-zero Algolia lane (after the 2026-08-08 03:15 + 2026-08-08 18:00 pulls); the 2026-08-10 + 2026-08-12 pulls had Algolia hits that were all string-collision noise. The Tavily web-search lane remains the better signal for vendor news.
- **Workflow lane:** **zero positive `openwiki` brand hits; zero `deepagents` hits.** `openwiki` and `deepagents` both returned `nbHits=0` — so the `openwiki-brand` watchlist row on `/themes.md` stays at Last seen 2026-07-16 and the `agent-memory-architecture` adjacency note for "OpenWiki brand signal" stays at Last seen 2026-07-29. The 27 unique top+best items include one **directly relevant** AI-agent item for the `agent-memory-architecture` theme — and six **adjacent, high-engagement AI-agent / inference / model-launch items** that touch Robin's broader agent-stack lane but stay adjacent-not-promoted under the brief's "low-engagement items are watchlist by default; promotion requires recurrence, strong engagement, or corroboration" rule **and** the `/themes.md` scope discipline ("open-weight AI never earns a `themes.md` row" + workflow-tooling priority LOW):
- **49286073** "Show HN: MCP Memory – Fast Agent Memory Using Google's OKF and SQLite FTS5" (46 pts, 22 cmts, top feed, by `fellowgeek`, https://github.com/fellowgeek/mcp-memory). Show HN for a Model Context Protocol (MCP) server that implements agent-memory using Google's **Open Knowledge Format (OKF)** as the frontmatter schema and SQLite FTS5 as the retrieval backend. **This is the first Hacker News item in the post-2026-07-29 sweep that explicitly references Google's OKF frontmatter** — the same OKF schema that OpenWiki uses ([`/themes/03-ai-agents.md`](/themes/03-ai-agents.md) § "Cross-agent memory substrate: OpenWiki", and the `okf_version: "0.1"` bundle-root frontmatter at [`/index.md`](/index.md)). The implementation is small (Show HN surfacing by a single author, `fellowgeek`) and engagement is moderate (46 pts) — under the established "Watchlist promotion requires recurrence, strong engagement, or corroboration" rule, 46-pt engagement sits in the corroborate-but-not-strongly-confirm band. **However** the item matches an existing topic (`agent-memory-architecture`) AND corroborates the OKF frontmatter pattern that OpenWiki has used since at least the 2026-07-16 launch (`langchain-ai/openwiki` HN `49058424`, 96 pts) — so it promotes `agent-memory-architecture` Last seen → 2026-08-13 (not a new theme row, just a Last-seen + Evidence anchor refresh on the existing row). The MCP-server framing also touches the MCP-revision watchlist that originally drove the 2026-07-28 [`hermes-agent-v0-19-quicksilver`](/themes.md) MCP-related Evidence cell. See `### agent-memory-architecture (refreshed this run)` below. **No new `/commitments.md` entry** — the item is a Show HN surfacing rather than a vendor-direct release, and Robin's OpenWiki already uses OKF (per `/index.md` front matter `okf_version: "0.1"`), so the synthesis layer does not need a new follow-up.
- **49285244** "DeepSeek Harness developer preview" (422 pts, 197 cmts, top+best, by `explosion-s`, https://deepseek.com/harness/en/). DeepSeek's developer preview for an "agent harness" — the orchestration layer over DeepSeek models for tool-using workflows. Strong topical fit to the AI-agent-inference lane and to the `MiniMax-m3` competitive-benchmarks strand (after the DeepSeek V4 Flash 0731 + DeepSeek V4 Pro 0813 + Tavily benchlm.ai comparison observations in this week-and-a-half). However, DeepSeek is an open-weight lab and the harness is an open-weight agent runtime, so per scope discipline ("open-weight AI never earns a `themes.md` row" + workflow-tooling priority LOW) the finding stays adjacent-not-promoted. **Third-party ANCHOR**: the 197-comment count exceeds the prior `49274600` DeepSeek V4 Pro 0813 (274 pts but 5 cmts) and `49214008` DeepSeek V4 Flash 0731 (494 pts); DeepSeek-side activity is real but still off-scope for themes.md. **Adjacent-not-promoted** per brief.
- **49281916** "Codex in ChatGPT desktop app for Linux is now in preview" (394 pts, 269 cmts, top+best, by `OXPHOS`, https://community.openai.com/t/codex-in-chatgpt-desktop-app-for-linux-is-now-in-preview/1390027). OpenAI community post announcing the **Codex surface in the ChatGPT Linux desktop app** is now in preview. **Strongest adjacents to Robin's Codex CLI fallback** — Codex is part of Robin's named-stack fallback (per [`/themes/03-ai-agents.md`](/themes/03-ai-agents.md) § "Codex CLI (fallback)") and the Linux desktop-app surface is a new Codex delivery channel that overlaps with Robin's Linux-using environment (WSL primary, plus the planned ITX 24h host — see `/commitments.md` `itx-24h-host-build-2026-q3`). However, (a) "Codex in ChatGPT desktop app" is a different surface from the "Codex CLI" fallback Robin currently references (CLI vs desktop-app channel), and (b) the established scope rule on Codex-wrapper/TUI changes keeps them adjacent-not-promoted unless there is a direct change to Hermes / MiniMax / OpenWiki / Codex CLI itself (per the 2026-07-27 / 2026-07-30 / 2026-08-03 / 2026-08-10 / 2026-08-12 traceability entries below). **Adjacent-not-promoted** per the scope rule. Flagged here because the 394-pt + 269-cmt engagement is high enough to surface in any HN digest even if the finding is not promoted. If Robin starts using the Codex-in-ChatGPT desktop app on Linux (instead of Codex CLI as fallback) on a future pull, the existing `Codex CLI (fallback)` section in [`/themes/03-ai-agents.md`](/themes/03-ai-agents.md) § "Adjacent workflow signals (HN / Tavily)" or its "Codex CLI adoption question" backlog item could be revisited.
- **49289112** "Gemini 3.7 Flash" (134 pts, 84 cmts, top feed, by `masont`, https://ai.google.dev/gemini-api/docs/models/gemini-3.7-flash). Google's Gemini 3.7 Flash release on the Gemini API. New model release; **not** on Robin's named daily-driver stack (Hermes / MiniMax-M3+H3 / OpenWiki / RAGFlow / NocoDB / SeaFile / Codex CLI). Adjacent to the LLM lane but off-scope per brief.
- **49285327** "Choosing an AI model: one prompt, 11 models, different results" (115 pts, 56 cmts, top feed, by `Philpax`, https://www.netlify.com/blog/one-prompt-11-models-very-different-results/). Netlify blog comparison of 11 LLMs on a single prompt. Strongest adjacents to the competitive-benchmarking lane already captured on the `MiniMax-m3` row (DeepSeek V4 vs MiniMax M3 strand) — but the Netlify blog is a marketing-influenced comparison, not a primary benchmark. Adjacent-not-promoted per scope. If a future Netlify blog benchmarks the 11 models including MiniMax-M3 specifically, that becomes a direct item on the `MiniMax-m3` row.
- **49285718** "I built a 500k-domain search engine for makers in a weekend for $10" (87 pts, 47 cmts, top feed, by `alexmorleyfinch`, https://alexmorleyfinch.github.io/marlin/history/v1/article/the_birth.html). Indie builder essay. Adjacent to the search-infra lane but not on Robin's stack. Off-scope per brief.
- **49286357** "We eliminated 1,400 CVEs in NanoClaw's container images" (53 pts, 37 cmts, top feed, by `jtbarker`, https://www.echo.ai/blog/echo-xnanoclaw-under-the-hood). Echo.ai blog post on container-image CVE remediation. Adjacent to the AI-agent-infra lane (NanoClaw is an AI-customer-facing agent runtime) but **not** on Robin's named stack. Off-scope per brief.
- **49288889** "Mistral OCR 4.1" (40 pts, 6 cmts, top feed, by `masont`, https://docs.mistral.ai/models/ocr-4-1). Mistral OCR 4.1 release. Adjacent to the OCR / document-understanding lane but not on Robin's named stack. Off-scope per brief.
- **49283063** "Launch HN: Bullet (YC S26) – A Faster Coding Agent" (17 pts, 19 cmts, top feed, by `sukeesh`, https://www.codewithbullet.com). YC S26 coding-agent launch. Adjacent to the AI-coding-agent lane but off-scope per the brief (not on Robin's named stack; the established scope rule keeps generic coding-agent launches adjacent-not-promoted per the 2026-07-27 / 2026-07-30 / 2026-08-03 / 2026-08-10 traceability entries).
- **49285770** "Ordinary Abundance" (80 pts, 28 cmts, top feed, by `mtwn`, https://ordinaryabundance.com/) — essay blog, off-scope.
- **49286485** "Kubernetes on Oxide" (91 pts, 30 cmts, top feed, by `mitchellh`, https://oxide.computer/blog/kubernetes-on-oxide) — Oxide blog, off-scope infra.
- **Other feed items not promoted:** 16 additional top/best items, all off-scope per the brief's scope filter — `49276574` "Delta" (645 pts, 237 cmts, best — Zed editor release, off-scope consumer IDE); `49286341` "Spaghettifying DRAM" (285 pts, 83 cmts, top — github drama on a code repo, off-scope); `49285982` "Gloomberb" (279 pts, 145 cmts, top — game, off-scope); `49284774` "Deutsche Bank becomes first foreign yuan clearing bank in Europe" (331 pts, 352 cmts, top — finance, off-scope); `49279928` "Principia Mathematica is modern and insightful" (259 pts, 137 cmts, best — essay, off-scope); `49286662` "I requested a copy of my data from McDonald's loyalty program" (135 pts, 157 cmts, top — privacy essay, off-scope); `49279638` "How art invented humanity" (36 pts, 5 cmts, top — humanities essay, off-scope); `49288296` "GoAccess" (7 pts, top — log analyzer, off-scope); `49279627` "The Indo-European Family Tree" (33 pts, 18 cmts, top — linguistics, off-scope); `49289512` "Choose Boring Technology (2015)" (8 pts, top — retro essay, off-scope); `49289465` "Donkey.bas is 45 Years Old – 131 line of Glory" (9 pts, 2 cmts, top — retro, off-scope); `49284697` "ATG (YC F25) Is Hiring Member of Technical Staff (Data Platform)" (1 pt — YC jobs, off-scope); `49289026` "Tocharian Online" (13 pts, top — linguistics, off-scope); `49288293` "We Have AI at Home Chapter 1: A Box of Scraps" (14 pts, 3 cmts, top — AI hardware essay, off-scope); `49286030` "Time to Move On: Querying Without Nulls and Bags" (40 pts, 10 cmts, top — arxiv database paper, off-scope). Engagement range 1–645 pts across the 16 items; none matches a Robin in-scope vendor / customer / application-domain / named-stack keyword.
- **Synthesis result:** **one theme Last-seen refresh, no commitment row, no open-question row, no personal-logistics item.** The `agent-memory-architecture` row on `/themes.md` refreshes Last seen → 2026-08-13, Evidence count → 16+, plus a new evidence anchor for the `49286073` MCP Memory / Google's OKF Show HN (`/themes.md` row description extended with the new evidence + a new Evidence cell anchor). All six flag-for-traceability items stay adjacent-not-promoted under the existing scope discipline. **Key call**: `49286073` MCP Memory using Google's OKF is the **first corroborating HN item on the OKF-frontmatter-as-agent-memory-schema pattern** since the 2026-07-16 `langchain-ai/openwiki` brand signal and the 2026-07-29 Tavily OKF-cluster pulls — it directly maps to OpenWiki's own OKF usage at `/index.md` and to the `agent-memory-architecture` watchlist that has been accumulating Karpathy LLM Wiki + OKF + MCP since 2026-07-26. The 46-pt engagement is moderate, but the **direct OKF reference + Show HN surfacing + MCP-server delivery channel** makes it a stronger trigger than the 2-pt `49156055` verginglabs Markdown-wiki post (2026-08-03) or the 1-pt `49159148` governed/approval-gated-memory post (2026-08-03) in the prior sweep. The MCP Memory implementation is on the same OKF + FTS5 pattern as a single-author Show HN, so it does not constitute a repeat observation on a high-engagement vendor release — but it does constitute the first HN-side corroboration of the OKF+memory schema that OpenWiki has been using. The `class-organoid-3d-bioprinting` row stays at Last seen 2026-08-12 (refreshed last pull); this run's workflow-lane item surfaces from the feed rather than from the Algolia query lane.
| Query lane | Queries run | Qualifying results |
| --- | --- | --- |
| Life sciences / vendors | `organoid`, `CelVivo`, `3Brain`, `Cellink` | 0 (all six queries returned `nbHits=0`) |
| AI workflow | `openwiki`, `deepagents` | 0 (both queries returned `nbHits=0`) |
Raw dump: `~/.openwiki/connectors/hackernews/raw/2026-08-13T18-00-02-662Z/hackernews-results.json`.
### agent-memory-architecture (refreshed this run)
- **49286073** — score 46 — top feed — "Show HN: MCP Memory – Fast Agent Memory Using Google's OKF and SQLite FTS5" by `fellowgeek` (https://github.com/fellowgeek/mcp-memory). Show HN for an MCP-server implementation that uses Google's **Open Knowledge Format (OKF)** as the canonical memory schema and SQLite FTS5 as the retrieval backend. Direct topical fit to the existing `agent-memory-architecture` theme (Last seen 2026-08-03, 15+ distinct sources per `/themes.md` row description) — and **the first Hacker News pull since 2026-07-16 to surface an item that explicitly names the OKF frontmatter format OpenWiki uses**. Engagement is 46 pts (moderate, above the 1-2-pt adjacent items in the 2026-08-03 sweep but below the 100+ pts threshold for high-engagement vendor releases) — enough to refresh Last seen on the existing row per the established pattern of "single-source mid-engagement + direct topical match + corroborates the existing pattern" (see the established rule for refreshing Last seen on the 2026-08-03 Hermes Agent v0.20.0 row at 1 pt + corroboration from the GitHub release tag), but not enough to launch a new theme row or commitment row. The MCP-server delivery channel aligns with the 2026-07-28 MCP-protocol revision (a related protocol-layer signal that already feeds the [`hermes-agent-v0-19-quicksilver`](/themes.md) row description). No follow-up commitment added; OpenWiki's own OKF usage at `/index.md` frontmatter (and at the per-concept frontmatter required by the OKF v0.1 schema) is already in place, so this item does not change Robin's workflow or trigger new documentation work.
### Flag-for-traceability: DeepSeek Harness, Codex on Linux desktop, Gemini 3.7 Flash, Mistral OCR — adjacent, off-scope
- **49285244** (422 pts top+best, 197 cmts) DeepSeek Harness developer preview — orchestration layer over DeepSeek models for tool-using workflows. Strongest DeepSeek-related signal of the pull (197-cmt engagement exceeds the prior DeepSeek V4 Pro 0813's 5 cmts), but DeepSeek is open-weight and the Harness is an open-weight agent runtime, so per scope discipline the finding stays adjacent-not-promoted. The DeepSeek-vs-MiniMax competitive lane is now 4 observations in 10 days (DeepSeek V4 Flash 0731 + Tavily benchlm.ai + DeepSeek V4 Pro 0813 + DeepSeek Harness developer preview) — if Hermes / MiniMax-side response (e.g. M3.5 release or new Headroom benchmark) on the same competitive pairing surfaces on a future pull, that becomes the trigger for an `MiniMax-m3-row-annex-competitive-benchmarks` description extension.
- **49281916** (394 pts top+best, 269 cmts) Codex in ChatGPT desktop app for Linux — new Codex delivery channel in the ChatGPT Linux desktop app. Codex is on Robin's stack as fallback (per [`/themes/03-ai-agents.md`](/themes/03-ai-agents.md) § "Codex CLI (fallback)") and the Linux desktop-app surface overlaps with Robin's Linux-using environment. However, "Codex in ChatGPT desktop app" is a different surface from "Codex CLI" (CLI vs desktop-app channel), and the established scope rule on Codex-wrapper/TUI changes keeps them adjacent-not-promoted unless there is a direct change to Hermes / MiniMax / OpenWiki / Codex CLI itself. **Adjacent-not-promoted** per scope.
- **49289112** (134 pts top, 84 cmts) Gemini 3.7 Flash — Google Gemini API release; off-scope.
- **49285327** (115 pts top, 56 cmts) Choosing an AI model: one prompt, 11 models, different results — Netlify blog comparison; off-scope.
- **49288889** (40 pts top, 6 cmts) Mistral OCR 4.1 — Mistral model release; off-scope.
- **49283063** (17 pts top, 19 cmts) Launch HN: Bullet (YC S26) – A Faster Coding Agent — YC S26 coding-agent launch; off-scope.
- **49286357** (53 pts top, 37 cmts) NanoClaw container CVE elimination — AI-agent-infra adjacent; off-scope.
## Update — 2026-08-12 18:00 UTC
- **Coverage:** 2 feeds (`top`, `best`), 6 search queries (the same simplified LLM-agent topic set from the source-specific instruction "Track organoid / CelVivo / 3Brain / Cellink / openwiki / AI agent stories": `organoid`, `CelVivo`, `3Brain`, `Cellink`, `openwiki`, `deepagents`). `top` feed returned 25 items, `best` feed returned 10; deduplication gave **30 unique items across both feeds** (5 overlap: `49271994` AI middle class software engineering, `49272832` Tailscale SQLite WAL-Reset bug, `49270953` 2026 Eclipse Webcams, `49273165` license plate warrant, `49274600` DeepSeek V4 Pro 0813). The 6-query lane produced the hit table below; the connector completed normally at `fetchedAt: 2026-08-12T18:00:30.849Z` (24-hour window).
- **Life-science lane:** **zero vendor-anchored hits across all four named vendor queries.** `CelVivo` returned `nbHits=0` (no stories indexed in 24h). `3Brain` returned 11 hits — one is a real organoid essay that was also surfaced by `organoid` (`49266068` "AI Is Dead. Organoids Are Alive", 5 pts, Wired — see "class-organoid-3d-bioprinting (refreshed this run)" below); the remaining 10 are string-collision noise on the substring "brain": `49263658` "The brain may be about to have its Ozempic moment" (114 pts, Economist — generic neuroscience essay on GLP-1-style brain drugs, outside Robin's 4 application domains); `49267390` "Dogs can tell when you're happy, sad or frustrated" (22 pts, Washington Post — consumer / veterinary neuroscience); `49272228` "Dog brains represent human facial expressions" (3 pts, Cell/iScience — veterinary); `49270499` "Big Tech Wants to Harvest Your Thoughts" (4 pts, Wired book excerpt — neurotech policy); `49271182` "Why I Play Board Games (and So Should You)" (4 pts, brainbaking.com — string hit on the "brainbaking" blog name); `49268526` "Brain scans show that dogs can distinguish various emotions in human faces" (1 pt, univie.ac.at — veterinary); `49269234` "New AI chip mimics the human brain's capacity for split-second motor control" (1 pt, LiveScience — brain-inspired computing chip, not 3Brain MEA / HD-MEA); `49269176` "My Terminal Workflow for Note-Taking, Data Engineering and Writing (Linux/macOS)" (1 pt, ssp.sh — "brain" tag slug, off-scope); `49273216` "Show HN: Kery – comments on your PR with a video of the feature working" (1 pt, github.com/Kery-HQ/Kery — `Kery-HQ` slug, off-scope). `Cellink` returned 4 hits — all string-collision noise on author-name / body-text fragments: `49273922` "Show HN: OJCP – an open protocol for agent-consumable job data" (8 pts, ojcp.dev — author `fraywing`, no relationship); `49275744` "Trump signs executive order calling for fewer childhood vaccines" (3 pts, CNBC — author `Fellw`, no relationship); `49265119` "Show HN: Supply-Wizard; Start Selling Data with an Easy Vendor Checklist" (3 pts, brickroad.network — string hit on "Checklist" / vendor terminology, no Cellink overlap); `49275608` "Development Mindset for AI Programming" (1 pt, no URL, author `kva`, no relationship). `organoid` returned 7 hits — **one real organoid essay** (`49266068` "AI Is Dead. Organoids Are Alive", 5 pts, Wired, by `Logans_Run`, https://www.wired.com/story/organoids-lab-grown-brains-neural-networks/ — see "class-organoid-3d-bioprinting (refreshed this run)" below) + six string-collision noise hits: `49275420` "Maintaining an organizational knowledge graph with an LLM and event sourcing" (2 pts, blog.arkency.com — "organizational" string); `49275020` "Ask HN: Responsible/trustable background check companies?" (4 pts, no URL — "organization" string); `49271578` "Show HN: Rekall – an Ableton Live project management companion application" (5 pts, rekall.machinelies.eu — author `iamsaitam`, "organization" string); `49270304` "Show HN: BentoDesk – Elegant Windows desktop organizer powered by Rust" (3 pts, github.com/ZRainbow1275/bentodesk — "organizer" string); `49266230` "Self-contained agents that self-organize" (3 pts, antigma.ai — "self-organize" string); `49264601` "Best Agent Gateways for Healthcare Organizations 2026" (1 pt, mintmcp.com — "Organizations" string). A title/URL regex sweep across the 30 unique top+best items also turned up zero organoid / 3D-bioprinting / CelVivo / 3Brain / Cellink / ClinoStar / ClinoReactor / MEA / spheroid / organ-on-chip items. The "no material HN footprint for the 4 vendor lines" finding still holds; the Tavily web-search lane remains the better signal for vendor news.