01Log
Archive
The full log, newest first.
macOS Debloat CLI
mac-os-debloat is a terminal utility that disables 296 non-essential macOS services to free up resources. It is notable for its SIP-safe approach, zero dependencies, and focus on reducing telemetry and system bloat.
SubnetDesk Remote Desktop
SubnetDesk is a fork of RustDesk focused exclusively on remote desktop within local networks and VPNs. It is notable for removing all public cloud, relay, and account dependencies to enhance privacy for homelab users.
Mufy Batch Export
mufy-batch-export is an open-source toolkit for exporting and archiving Mufy chat data into formats like Markdown, JSON, and EPUB. It is notable for enabling data portability and offline reading from a proprietary platform.
Recursive self-improvement of AI research agents
Paper on agents that recursively improve their own research workflows — the loop where an agent's output becomes its next input. Relevant to anyone running autonomous research or coding loops.
RoboFollow: Unveiling the Instruction Following Mirage in Embodied Agents
CoRL 2026 paper with a simulation benchmark showing embodied agents claim instruction following but fail under test — a robustness gap that mirrors what operators see in coding agents.
Just-in-Time Memory: Learning to Curate Task-Adaptive Memory for LLM Agents
Proposes read-time, task-adaptive memory curation instead of the write-time distillation most agent frameworks default to — a fix for agents that drown in stale summaries.
PACT: From Credit Assignment to Critic Alignment
Formalizes token-level credit assignment via three conditions — directly relevant to RL-trained agents where sparse rewards make critic alignment the bottleneck.
How we made claude.ai 3x faster in two weeks
Anthropic used Claude to measure and optimize claude.ai, cutting latency 3x in two weeks — a case study in self-improving agent workflows, with HN discussion on reward hacking in kernel optimization.
Advancing Private AI Compute with secure, server-side memory
DeepMind details secure, server-side memory for private AI compute — relevant if you're running agents on sensitive data and need enclave-backed memory.
Devin September 23, 2026
Devin Sept 23 — parent sessions now show the child session tree; permission waits resume immediately on response; autoreview triggers become composable rules (account/SCM org/repo/prefix/member + filters). Capability picker names models: Ultra = Fable 5.1 + GPT-6 Astra.
Devin 1 more releases
Devin CLI stable v3000.11 — run Devin Cloud sessions from the terminal (`devin --cloud`), handoff/pickup cloud PR branches locally, `devin ssh` into the cloud VM, ACP model/thinking/speed controls, and OTel export. v3000.11.3 fixes tenant-aware session links on dedicated tenants.
NanoClaw 2.4.0
NanoClaw 2.4.0 — credential gateways via skills (Iron Proxy new), install-wide/per-group model and speed controls, Mattermost channel, reworked OpenCode provider. Breaking: custom instruction sources must move to the new module — CLAUDE.md @-imports outside the project dir were silently dropped, so most instructions never reached the model. Defaults shift to Opus 5.5 and gpt-6-astra.
Ember 1
Fireworks Research's specialized reasoning model built on Kimi K3, with 1M context and $3/$15 pricing.
Holo4 27B
Open image-text-to-text generalist for non-commercial multimodal work.
Holo4 35B A3B
Apache-licensed sibling with 35B total params for commercial multimodal use.
Ringg’s AI agents resolve up to 65% of customer calls with OpenAI
Ringg reports its OpenAI-powered agents resolve up to 65% of customer calls autonomously — a concrete production benchmark for customer-service agent deployments.
Can Agents Design Better Chips with a Higher Level Abstraction?
Asks whether agents can design better chips when given higher-level abstractions — early evidence on agent capability in hardware design workflows.
When LLM Agents Fail to Read the Room: ReAdapt for Relational Social Reasoning
Proposes ReAdapt, a method for improving LLM agents' relational social reasoning — targeting failures where agents miss conversational context and social cues.
The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks
Introduces a benchmark for 'taste' in long-horizon agent tasks — measuring whether agents produce outputs that match human quality preferences, not just complete the task.
SF October 14th: A Birds of a Feather Session on Agentic Engineering
Simon Willison and Jesse Vincent host a Birds of a Feather on agentic engineering in SF Oct 14 — an evening for practitioners to compare notes on building and operating agents.
Infinite Canvas Tutorial
A tutorial for building infinite canvas applications using WebGL and WebGPU, covering high scalability, zooming, and direct manipulation. It is notable for its technical depth on a modern topic and is freshly maintained.
Visualization Collection
A collection of over a hundred frontend visual effect examples, including CSS animations, Canvas, and Three.js 3D. It serves as a useful reference for visual design but is not a standalone tool.
Inkstone Markdown Notebook
Inkstone is a self-hosted Markdown notebook application that runs entirely on Cloudflare Workers. It is notable for its novel serverless architecture and momentum in the self-hosted knowledge management space.
Farm Build Tool
Farm is an extremely fast, Vite-compatible web build tool written in Rust, designed to avoid Vite's performance issues on large projects. It is notable for bringing Rust-based build speeds to Vite-compatible workflows and is actively maintained.
Prehistoric Animal Museum
A free, open-source, bilingual 3D museum where young children can explore prehistoric animals with narration and a parent guide. It is notable as an educational 3D project, though its audience is niche.
PDFX PDF Processor
PDFx is a desktop application that provides a free-floating 2D canvas for viewing, rearranging, and combining multiple PDF files and images simultaneously. It is notable for its novel visual approach to multi-document PDF handling.
SaaS Boilerplate
SaaS Boilerplate is an open-source starter stack using React, Django, and AWS to build SaaS products faster. It is a solid, full-stack boilerplate but not specifically novel or frontend-focused.
VideoGen-Agent
A framework for reinforcing agents specifically for video generation tasks.
Agensh
A paper proposing a self-organized multi-agent harness to scale organizational intelligence to 1,024 agents without central orchestration.
Emergent Collusion in Long-Horizon LLM Agent Interaction
A paper studying if agents jointly violate instructions during long-term interactions.
The new CC
50 points on HN; users want this for consolidating fragmented school schedules and family logistics.
Radical Numerics
Analysis on using biological chain-of-thought and multimodal perception to counter AI-driven bio-security threats.
Devin CLI v3000.11.1
Devin CLI v3000.11.1 adds cloud session control from the terminal — `devin --cloud`, `/handoff`/`/pickup` to pull a cloud PR branch local, and `devin ssh` into the VM. Also OpenTelemetry export and ACP model selection.
Anthropic SDK 0.128.0
Anthropic TS SDK 0.128.0 adds claude-opus-5-5 support, inline tool definitions, and MCP tool-list pinning (beta); `addTools()` now takes effect immediately.
OpenHands 1.22.0
OpenHands 1.22.0 lets cloud backends test remote MCP servers via the app server and opens Git Sync to org admins; also fixes preserving sibling MCP servers when saving to cloud.
GPT-6 Sol
Cost-efficient high-end tier in the GPT-6 series, positioned below Astra and above Luna.
Aion 3.5
Multi-model roleplaying and storytelling system built on GLM, using collaborative generation across specialized models.
Command A+
Cohere's flagship for enterprise agentic workflows with native tool calling and strict schemas.
Claude Opus 5.5
Anthropic's flagship for demanding reasoning, coding, and long-horizon agentic work, succeeding Opus 5.
Rho-RoboEval
Microsoft's 5.2B model for robotics evaluation.