Run notes · 2026-07-07 · resource-discovery-run1

Resource Discovery Report — 2026-07-07

Public notes rendered from loops/log/2026-07-07-resource-discovery-run1.md.

Resource Discovery Report — 2026-07-07

1. What changed

  • Swept committed sources (resources/SOURCES.md) plus rotating RSS/url/repo sources against 40 entries already in the library.
  • Run #1 (reconstructed for durable log). Rotating exploratory queries: arXiv cs.AI (agent loop / agentic) and arXiv cs.SE (autonomous agent software engineering) via RSS; HN frontpage as a lead list only. Recorded for next-run dedupe and rotation.
  • 31 new candidate(s) after dedupe against the library and loop state:
  • tencent/Hy3

- Source type: rss - Summary: tencent/Hy3 New Apache 2.0 licensed model from Tencent in China: Hy3 is a 295B-parameter Mixture-of-Experts (MoE) model with 21B active parameters and 3.8B MTP layer parameters, developed by the Tencent Hy Team. Following the Hy3 Preview launch in late April, we gathered feedback from 50+ products and scaled up post-training with higher quality data. Today, we introduce Hy3, which outperforms similar-size models and rivals flagship open-source models with 2-5x parameters. It also shows significant gains in utility across various products and productivity tasks. The full-sized model is 598GB on

- Source type: rss - Summary: Release: sqlite-utils 4.0rc3 I hoped to release sqlite-utils 4.0 stable this weekend, but as I worked through the backlog of issues and PRs with a combination of Claude Fable 5 and GPT-5.5 the changelog since rc2 kept getting bigger . The biggest new feature is support for introspecting and creating compound foreign keys - a feature that involves a subtle breaking change to table.foreign_keys and hence needed to land for the 4.0 stable release. sqlite-utils also now follows SQLite's convention for case insensitive column names, which turned out to touch a bunch of different places at once . Ta

- Source type: rss - Summary: I wrote about the sqlite-utils 4.0rc1 release a couple of weeks ago. Since we only have Claude Fable on our Max subscriptions for a few more days, I decided to see if it could help me get to a 4.0 stable release that I felt truly comfortable about, since I try to keep to SemVer and like my incompatible major versions to be as rare as possible. I started with this prompt, in Claude Code for web on my iPhone: Final review before shipping a stable 4.0 release - very important to spot any last minute things that would be a breaking change if we fix them later Here's that initial report it created

- Source type: rss - Summary: Release: sqlite-utils 4.0rc2 See sqlite-utils 4.0rc2, mostly written by Claude Fable (for about $149.25) .

- Source type: rss - Summary: Building a World Map with only 500 bytes Iwo Kadziela (assisted by Codex) figured out a way to generate a credible ASCII world map using 445 bytes of data: The key trick is to use deflate compression, which is then wired together using this neat snippet of JavaScript. I didn't know you could use fetch() with data: URIs like this: fetch('data:;base64,1ZpLsgIxCEXnrM...==').then( r => r.body.pipeThrough(new DecompressionStream('deflate-raw')) ).then( s => new Response(s).text() ).then( t => b.innerHTML = '<pre style=font-size:.65vw>' + t ) Via Hacker News Tags: ascii-art , data-urls , javascript

- Source type: rss - Summary: AI gets good at anything with an answer key. Your career is everything that doesn't have one.

- Source type: rss - Summary: The action in agentic engineering has moved from prompting to operating. Autonomy isn't one ladder, it's two axes (agency and orchestration) and six levels you move between per task. The real question is what level a task deserves, and what verification makes that level defensible.

- Source type: rss - Summary: I co-wrote a Google whitepaper about how AI is changing the software lifecycle. I'm not going to summarize the whole thing. Instead, here are the handful of ideas in it I think actually matter, plus six figures you're welcome to reuse.

- Source type: rss - Summary: Coding agents are extraordinarily good now, and getting better fast. The interesting consequence is that the hard part of engineering moved from writing code to deciding whether to trust it, which makes review the most leveraged skill in software right now. How you approach it depends enormously on who you are: a solo developer with no users and a team maintaining a ten-year-old application are not solving the same problem.

- Source type: url - Summary: Harrison Chase | Substack <meta data-rh="true" name="twitter:image" content="https://substackcdn.com/image/fetch/$s_!vdHI!,f_auto,q_auto:best,fl_progressive:steep/https%3A%2F%2Fsubstack.com%2Fapi%2Fv1%2

- Source type: url - Summary: METR { "@context" : "https://schema.org", "@type" : "WebSite", "name" : "METR", "url" : "https://metr.org/" }

- Source type: url - Summary: Build powerful multi-agent systems with Agent Development Kit (ADK)

- Source type: url - Summary: Redirecting to LangGraph Documentation var anchor=window.location.hash.substr(1);location.href="https://docs.langchain.com/oss/python/langgraph/overview"+(anchor?"#"+anchor:"") Documentation has moved The LangGraph documentation has moved to docs.langchain.com . Redirecting you now...

- Source type: url - Summary: Build collaborative AI agents, crews, and flows — production ready from day one.

- Source type: github_repo - Summary: τ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains - sierra-research/tau2-bench

- Source type: github_repo - Summary: Contribute to openai/simple-evals development by creating an account on GitHub.

- Source type: rss - Summary: arXiv:2607.02542v1 Announce Type: new Abstract: General-purpose embodied agents must understand multimodal instructions, anticipate how their environment will evolve, and produce precise control actions over extended horizons. Existing approaches typically specialize in visual-language reasoning, video-based world modeling, or action generation, while cascaded pipelines that first synthesize future observations and then infer actions can introduce interface bottlenecks and compound prediction errors. We present iFLYTEK-Embodied-Omni, a unified multimodal foundation model that jointly models vi

- Source type: rss - Summary: arXiv:2607.02672v1 Announce Type: new Abstract: Local pairwise comparisons are a standard tool for learning how people want decision rules to work, e.g., in participatory design or alignment. However, their use builds in two strong assumptions: that local comparisons are sufficient evidence about how a person wants an automated decision rule to behave, and that people can always answer those comparisons decisively. We investigate how these assumptions may be compromised under internal pluralism: the idea that an individual evaluates decision rules according to multiple authoritative priorities

- Source type: rss - Summary: arXiv:2607.02686v1 Announce Type: new Abstract: Reinforcement learning agents operating under partial observability must act on incomplete information, making them natural candidates for guidance from small language models (SLMs) that carry broad reasoning priors. Yet integrating SLM guidance into this setting has proven difficult: across all test environments, vanilla uncertainty-gated approaches achieve an overwrite rate at or near zero, meaning the SLM almost never contributes an independent action. We trace this failure to the bare egocentric prompt, which provides insufficient context for

- Source type: rss - Summary: arXiv:2607.02771v1 Announce Type: new Abstract: Leadership computing facilities steward large-scale scientific datasets that routinely require substantial transformation before serving as AI training data. However, no existing framework fully unifies automated transformation, readiness assessment, provenance tracking, and agent-native deployment. We present REDI, an open-source framework that addresses this gap through a unified five-stage pipeline (ingest, preprocess, transform, structure, and output) with per-stage instrumentation for reproducibility and deployment as an agent-callable skill

- Source type: rss - Summary: arXiv:2607.02807v1 Announce Type: new Abstract: Long-running coding agents such as autoresearch can persistently discover optimizations for open-ended problems. However, they tend to converge onto a single high-level approach, then proceed with low-level edits while missing other superior approaches to the problem. We hypothesize two harness-level design choices contribute to this behavior: accumulating context in a single long-running agent and only exposing a single program state to edit. We introduce SwarmResearch, an orchestrator-subagent harness in which a Shepherd Agent uses global conte

- Source type: rss - Summary: arXiv:2607.02577v1 Announce Type: new Abstract: Tool-calling benchmarks are increasingly used to rank language-model agents, yet their scores are often treated as ground truth without validating the evaluators themselves. We present a systematic validity and reproducibility audit of four major tool-calling benchmark families: BFCL v4, {\tau}2-Bench, LiveMCPBench, and MCP-Atlas. Across 496 expert-reviewed benchmark tasks, we find 92 evaluator-human disagreements, corresponding to an 18.5% misalignment rate. The failures are not isolated annotation mistakes: deterministic benchmarks exhibit brit

- Source type: rss - Summary: arXiv:2607.02579v1 Announce Type: new Abstract: Long-lived language agents increasingly write reusable memories from their own execution traces. The key safety question is not only what agents should remember, but when they should refuse to write memory at all. Repeated observations across agents are not necessarily independent evidence: the same claim may be copied from a shared source, induced by a shared prompt, stale under a new environment, or valid only in a narrower scope. We study this failure mode as a memory writepath governance problem. We introduce GovMem as a conservative diagnost

- Source type: rss - Summary: arXiv:2607.02583v1 Announce Type: new Abstract: A growing body of work combines large language models (LLMs) with classical optimizers for software engineering (SE) configuration tasks. Often, the classical optimizer is in charge: it owns the search loop and calls the LLM only to assist in subroutines (e.g. to warm-start the first generation, propose a mutation, or stand in as a surrogate). We report that there is much value in the reverse approach: seeding an LLM with the results from a cheap classical learner. We call this method SNAP2. Applied to over 100 SE tasks, it is the single best of

- Source type: rss - Summary: arXiv:2607.02587v1 Announce Type: new Abstract: Model cards quote trust-benchmark scores without recording when they were measured, and the same number is routinely carried across successive checkpoints of one release line as if the model behind it had not shifted. We test whether it has shifted by auditing four open-source release lines, Yi, Qwen, Mistral, and Gemma, at three successive generations each, on a fixed basket of trust benchmarks under multiple prompt templates. Mean absolute adjacent-generation drift lands well above an independence-based no-drift reference null, and the gap pers

- Source type: rss - Summary: arXiv:2607.02590v1 Announce Type: new Abstract: Transforming static research papers into dynamic media such as posters, slides, and videos is essential for effective dissemination but remains a labor-intensive challenge. Existing automated approaches often treat these formats in isolation and consequently fail to maintain semantic consistency across the entire presentation suite. We address this fragmentation by formalizing the task of unified presentation suite generation and proposing $\textbf{OmniPresent}$ to orchestrate the creation of coherent deliverables. Our framework adopts a renderab

- Source type: rss - Summary: Article URL: https://jobs.ashbyhq.com/lago Comments URL: https://news.ycombinator.com/item?id=48814509 Points: 0 # Comments: 0

- Source type: rss - Summary: Article URL: https://bradleywoolf.com/links-1/sequencing-my-own-dna-at-home Comments URL: https://news.ycombinator.com/item?id=48812156 Points: 195 # Comments: 69

- Source type: rss - Summary: Article URL: https://spectrum.ieee.org/small-language-models-ai-pharmaceuticals Comments URL: https://news.ycombinator.com/item?id=48812055 Points: 126 # Comments: 41

- Source type: rss - Summary: Article URL: https://blog.cr.yp.to/20260706-fairness.html Comments URL: https://news.ycombinator.com/item?id=48811887 Points: 100 # Comments: 84

- Source type: rss - Summary: Article URL: https://ternlight-demo.vercel.app/ Comments URL: https://news.ycombinator.com/item?id=48811644 Points: 213 # Comments: 49

  • Each candidate above links to the fetched URL or repo-relative file path.
  • Library dedupe source: /persisted/repos/loopmaster.ai/.worktrees/t_1266994d/resources/resources.yaml
  • Loop state file: /tmp/loopmaster-rd-state-run1-recon.json
  • Score each candidate against resources/RUBRIC.md (five dimensions, 0–2 each).
  • Open a review-only proposal (PR or kanban card) with ready-to-merge YAML for candidates scoring ≥7 with no dimension at 0.
  • Record at least one rejected candidate in the proposal to prove the rubric has teeth.
  • Do not merge proposed entries into resources/resources.yaml without human approval.

4. Confidence and risk

  • Confidence: medium for fetch/dedupe; rubric scoring and inclusion remain a human-or-reviewer judgment call.
  • Risk: source pages can change after fetch; candidates are proposals, not inclusions. Re-score on merge if the content has drifted.

5. Autonomy / approval status

  • Autonomous action taken: wrote this report and updated dedupe state.
  • Human approval required before merging proposed entries into resources/resources.yaml, public publishing, product direction changes, or credential changes.

Fetch errors

  • Anthropic news: HTTP Error 404: Not Found
  • LangChain blog: not well-formed (invalid token): line 1, column 2326
  • Andrew Ng / DeepLearning.AI The Batch: HTTP Error 404: Not Found
  • Eugene Yan: HTTP Error 404: Not Found
  • OpenAI research index: HTTP Error 403: Forbidden