The dispatch

Recent updates.

Everything that shipped in AI this cycle, tested and boiled down for builders. Tap any to open the source.

All updates

80 updates · updated Jul 20
Models Jul 18 · 3-min read

GPT-5.6 closes a 30-year convex-optimization gap from a single prompt

An independent reproduction (HN front page Jul 18, press Jul 19) shows GPT-5.6 deriving a result that had stood open for three decades from one prompt. The takeaway for builders: frontier models can now do real research-grade math, so treat them as collaborators on hard analytical problems, not just code and prose.

Models Jul 16 · 3-min read

Moonshot ships Kimi K3: 2.8T open MoE, Opus-4.8-class at Sonnet-5 pricing

A 2.8-trillion-param open mixture-of-experts with a 1M-context window, benchmarking near Opus 4.8 while priced like Sonnet 5. Weights land Jul 27. Already routable via Vercel AI Gateway, so A/B it against your current frontier model on cost-per-task before the weights even drop.

Models Jul 19 · 2-min read

Alibaba previews Qwen3.8-Max: 2.4T multimodal, OpenAI- and Anthropic-compatible

Days after Kimi K3, Alibaba's 2.4-trillion-param flagship goes live in preview via Token Plan, Qoder, and QoderWork at 10% of standard pricing, with open weights promised soon. It speaks both OpenAI and Anthropic API protocols, so you can A/B it without restructuring your stack. Active-param count and independent benchmarks are still undisclosed, so verify before trusting the second-only-to-Fable-5 claim.

Copilots Jul 19 · 2-min read

Claude Code v2.1.215 stops auto-running /verify and /code-review

Those two commands are now opt-in instead of firing automatically at the end of a turn. If you relied on the auto-behavior, wire /verify and /code-review back into your own hooks or run them manually.

Copilots Jul 18 · 2-min read

Claude Code v2.1.214 patches permission bypasses and adds an EndConversation tool

Closes holes where Edit globs and long or quoted bash could dodge permission prompts, plus a new EndConversation tool that lets the agent cleanly close a session. Upgrade if you run Claude Code in any semi-autonomous mode.

Copilots Jul 18 · 1-min read

Codex CLI 0.144.6 corrects the GPT-5.6 context window to 272K

The Sol, Terra, and Luna tiers were mislabeled; the real window is 272K tokens. Update so your context-budgeting and truncation logic uses the right ceiling.

Security Jul 17 · 1-min read

Cloudflare emergency WAF release virtual-patches live RCE and SQLi

New managed rules block active remote-code-execution and SQL-injection exploits in popular WordPress frameworks before you can patch upstream. If you front any WordPress on Cloudflare, confirm the managed ruleset is enabled.

Automation Jul 17 · 2-min read

Neon promotes Object Storage, Functions, and AI Gateway to public beta

Plus SDK 1.1 and a new neon diff for schema comparison. Postgres, blob storage, serverless functions, and model routing now live behind one Neon project. Try it if you want to collapse your agent backend onto a single provider.

OpenSource Jul 17 · 2-min read

Capital One open-sources VulnHunter: agentic vuln-hunting as Claude Code skills

A packaged set of Claude Code skills that autonomously hunt for vulnerabilities in a codebase. Drop it into your repo to run agent-driven security passes on in-flight changes.

Copilots Jul 17 · 2-min read

Cursor's agent goes plan-first, multi-repo, and cross-channel

The agent now proposes a plan before editing, spans multiple repos in one task, and picks work up across channels. Review the plan gate to keep large multi-repo edits under control.

Copilots Jul 16 · 2-min read

Cline v4.0.9 cuts install size ~640MB to 285MB and adds Kimi K3

A Jul 19 follow-up moves heavy peer deps to optional, roughly halving the footprint, and Kimi K3 is now selectable in the CLI. Reinstall for a lighter agent and a cheaper frontier option.

Automation Jul 16 · 1-min read

Railway plugin lands in the xAI Grok Build marketplace, plus a new railway cdn CLI

Deploy and manage Railway services from inside Grok Build, and push edge caching from the terminal with railway cdn. Handy if you drive infra from an agentic CLI.

Models Jul 16 · 1-min read

Vercel AI Gateway adds Kimi K3 and GLM 5.2; Sandbox egress now free

Two new open frontier models one gateway call away, plus free outbound traffic from Vercel Sandbox and Slack agents in the Chat SDK. Swap models via config with no new SDK.

Copilots Jul 16 · 1-min read

Lovable adds workspace identity reuse plus Apollo.io and Wiz connectors

Share one identity across a workspace and pull data from Apollo and Wiz into builds. Reuse auth instead of re-connecting per project.

Copilots Jul 16 · 2-min read

GitHub Copilot CLI v1.0.72 pre-releases add multi-turn subagents and plugin management

Subagents can now hold a multi-turn conversation, and a preview plugin manager lets you add and remove capabilities. Try it for delegated, longer-horizon agent tasks.

Agents Jul 19 · 1-min read

Devin Desktop v3.5.17: instant worktree-backed sessions and per-session network policy

Each session gets its own git worktree so parallel work does not collide, and you can scope network access per session. Good for running several isolated agent tasks at once.

OpenSource Jul 19 · 2-min read

vercel-labs/deepsec breaks out on HN: a security harness for coding agents

A ~6k-star harness that wraps coding agents with security checks. Part of the wider agentic code-security wave alongside VulnHunter and Strix; worth studying if you run agents against production code.

OpenSource Jul 16 · 1-min read

xhluca/agent-talk hits the HN front page: a protocol for coding agents to collaborate

An open protocol that lets multiple coding agents coordinate on the same task. A reference to watch as multi-agent workflows go mainstream.

OpenSource Jul 19 · 1-min read

Show HN: hyperlogue/r3 reviews AI-generated code locally and feeds notes back to the agent

Runs a local review pass on agent output, then hands the findings back to the agent to fix. A concrete answer to the generation-versus-verification bottleneck.

OpenSource Jul 19 · 1-min read

Show HN: AdamACE9/bothread is a local MCP room for collision-free multi-agent co-editing

A local MCP server that coordinates several agents editing the same files without clobbering each other. Useful scaffolding for parallel agent workflows.

Automation Jul 16 · 1-min read

Cloudflare adds wrangler flagship to manage feature flags from the CLI

Toggle feature flags from CLI, CI, or an agent without a redeploy. Wire it into automation to flip behavior without a full deploy cycle.

Launch Jul 18 · 1-min read

ZooData launches: a data layer that turns any URL into agent-ready JSON

Point it at a URL and get structured, agent-ready JSON, exposed through an MCP server (~625 upvotes on Product Hunt). A quick way to feed live web data into agents.

Automation Jul 17 · 1-min read

Cloudflare Email adds a sent-email preview in the Activity log

See the rendered body of routed messages straight from the Activity log. Faster debugging for email-routing flows.

Copilots Jul 17 · 3-min read

Claude Code ships v2.1.208-2.1.213 with agent-safety controls

Session-wide WebSearch caps, per-session subagent spawn limits, auto-backgrounding of long MCP calls, /fork into background sessions, and a screen-reader mode. 2.1.212 fixed a plan-mode bug that ran file-modifying Bash without permission. Upgrade to 2.1.213 and use the new caps to bound runaway agent runs.

Models Jul 9 · 3-min read

OpenAI GPT-5.6 family lands: Sol, Terra, Luna

Frontier (Sol), balanced (Terra), and high-volume (Luna) tiers, plus Programmatic Tool Calling, explicit prompt-cache controls, persisted reasoning, and beta multi-agent orchestration in the Responses API. Route by tier and adopt Programmatic Tool Calling to cut agent-loop cost.

Models Jul 16 · 2-min read

xAI Grok 4.5 released, trained alongside Cursor for coding

xAI's strongest coding/agentic model at $2/1M in and $6/1M out, live in Cursor on all plans day one. Benchmark it for agentic coding before defaulting to it.

OpenSource Jul 15 · 2-min read

xAI open-sources the Grok Build coding CLI (and a privacy scare)

A full agentic coding harness you can fork and study. But a wire-level analysis found the CLI uploading entire tracked git repos to a GCS bucket, forcing a public deletion promise. Great reference harness, and a cautionary tale on agent data exfiltration.

Models Jul 15 · 3-min read

Thinking Machines drops Inkling: ~1T-param open multimodal model

An open-weight ~1T model that natively ingests text, image, and audio, with day-0 support in transformers, SGLang, and llama.cpp, plus routing on Vercel AI Gateway. An NVFP4 build (~600GB VRAM) or llama.cpp quants make it self-hostable; try HF Inference Providers first.

Agents Jul 7 · 2-min read

Google adds managed agents to the Gemini API: background tasks + remote MCP

Agents run async server-side (background: true) and call remote MCP servers. Offload long-horizon Gemini work to the background and reach private data via remote MCP instead of local shims.

Copilots Jul 14 · 2-min read

GitHub Copilot adds /security-review slash command (public preview)

Inline scan of in-flight changes for injection, XSS, path traversal, and weak crypto, with severity and confidence scores. Available on Free and Pro too. Run it on a diff before you open the PR.

Voices Jul 6 · 2-min read

OpenAI GPT-Realtime-2.1 voice model (plus mini)

Better interruption handling and alphanumeric recognition for production voice agents. Swap to gpt-realtime-2.1 (or the cheaper mini) for voice input, overlapping speech, or spelling of codes and emails.

Copilots Jul 16 · 1-min read

Codex CLI 0.144.5 widens dangerous-command detection

Blocks more destructive shell forms (expanded rm variants) before execution in autonomous modes. If you run Codex at higher autonomy, upgrade, though these builds are pre-release.

Copilots Jul 14 · 2-min read

GitHub Copilot for JetBrains expands BYOK

Point at any OpenAI-compatible endpoint with your own key, customize Claude agent providers, and run local sandboxing (preview). Bring a self-hosted model into JetBrains Copilot and keep it local.

Copilots Jul 16 · 2-min read

Copilot in Visual Studio: MCP trust layer + C++ modernization agent GA

Startup-time fingerprinting of MCP server config against a trusted baseline (a supply-chain guardrail), plus GA of automated MSVC upgrades. Verify MCP servers haven't been tampered with.

Automation Jul 15 · 1-min read

Cloudflare Internal DNS now GA

Public and private DNS unified with shared policy. Consolidate split-horizon DNS for agent and backend flows behind one set of internal resolvers.

Agents Jul 13 · 1-min read

Cloudflare Agents can answer MCP elicitation requests

Agents handle mid-flow form and URL user input over MCP. Build agents that pause to collect structured input and then resume.

Models Jul 16 · 1-min read

Vercel AI Gateway adds Kimi K3

The newer Kimi frontier model is one gateway call away, with no separate Moonshot integration. A/B it via gateway config, no new SDK.

Voices Jul 9 · 2-min read

Deepgram Voice Agent: barge-in interrupts + auto latency reports

Inject a message that interrupts an in-progress turn, plus full LLM, TTS, and end-to-end latency breakdowns emitted automatically. Build barge-in voice UX and diagnose latency without extra instrumentation.

Copilots Jul 10 · 2-min read

Replit Agent: buy domains in-app, Excalidraw connector, editor permissions

The Agent creates and edits diagrams without credential setup, and non-owner editors can answer the Agent's routine questions. Delegate Agent Q&A to teammates and wire domains in-app.

Automation Jul 10 · 1-min read

Neon CLI gains neon api + self-expiring MCP branches

The CLI hits any Platform API route, and MCP branch create accepts expiresAt. Spin up auto-cleanup DB branches from agents or CI.

Automation Jul 10 · 1-min read

Railway Agent lands in Slack and Discord

Deploy and manage infra from chat, plus feature flags and CLI spend caps. Trigger infra ops from chat and cap agent or CLI spend.

Voices Jul 7 · 1-min read

Cohere open-sources Transcribe Arabic ASR

Open-weights Arabic ASR that beat Whisper in 96% of human comparisons. Evaluate it as a Whisper replacement for Arabic audio.

Automation Jul 17 · 1-min read

Supabase ships Unified Logs across the platform

One log pane across Edge Functions, DB, and Realtime. Trace cross-service issues in a single view.

Models Jul 9 · 3-min read

Meta ships Muse Spark 1.1 and opens its first paid proprietary API

A 1M-context multimodal agentic model on a new OpenAI-compatible Meta Model API (US preview), reported near $1.25/$4.25 per 1M with $20 free credits. Point your OpenAI SDK base URL at it to A/B test.

Copilots Jul 8 · 3-min read

GitHub Copilot in VS Code gets parallel agent sessions and per-session cost

Run multiple agent sessions side by side with full-session cost visibility, plus Marketplace model-provider discovery and a unified context and reasoning picker.

Voices Jun 29 · 2-min read

xAI Grok voice models land on Vercel AI Gateway

grok-voice-think-fast-1.0 realtime plus grok-tts and grok-stt are callable through AI Gateway on AI SDK 7, with no separate xAI key and built-in spend controls.

Copilots Jul 10 · 2-min read

Cursor 3.11 adds parallel side chats, transcript search, and agent hooks

The /side command opens a durable side chat for tangents without polluting the main thread, Cmd+K searches agent transcripts, and new beforeSubmitPrompt, afterAgentResponse, and afterAgentThought hooks let teams instrument agents.

Automation Jul 9 · 2-min read

Supabase Realtime Broadcast adds binary payloads

Broadcast now sends and receives ArrayBuffer across clients, REST, and the DB (auto in supabase-js 2.91.0 and up), cutting encoding overhead for telemetry and live-frame streaming. Watch for silent drops to older clients.

Voices Jul 1 · 2-min read

Deepgram Voice Agent API adds managed Claude Sonnet 5

A stronger reasoning LLM inside Voice Agent API with no separate Anthropic key or orchestration. The claude-sonnet-4-20250514 snapshot is deprecated and now errors, so audit any pinned configs.

Copilots Jul 3 · 2-min read

Warp adds native Antigravity CLI agent and custom model routers

Native agy CLI agent support plus user-defined model routers turn the terminal into a multi-model routing hub. Route cheap tasks to small models and hard ones to frontier models automatically.

Copilots Jul 8 · 2-min read

Zed 1.10 makes llama.cpp a first-class model provider

Point Zed's agent at a local llama.cpp model for fully offline, private agentic coding. Note that format-on-save is now off by default.

Copilots Jul 8 · 2-min read

OpenAI Codex 0.143.0: remote plugins default, system proxy, MCP tool search

Remote plugins are on by default for a bigger catalog, and new macOS and Windows system-proxy support finally lets corporate users behind proxies run the Codex CLI cleanly.

Copilots Jul 9 · 2-min read

OpenAI Codex 0.144.0 adds a new "writes" approval mode

A safer middle ground that approves edits but not arbitrary commands, plus MCP auth without opt-in. Useful for teams that want agents to edit files but not execute.

Copilots Jul 7 · 2-min read

Lovable adds Design Guidance, a GitHub connector, and Project Monitoring

Design Guidance generates three design directions before building to fight generic AI aesthetics, and a new GitHub API connector wires repos, issues, and PRs into issue-aware builds.

Voices Jul 2 · 1-min read

Deepgram Flux adds word-level timestamps

Flux conversational STT now returns per-word start and end times for precise captions, highlighting, and barge-in timing. Bump the Python SDK to 5.3.4 or later if on 5.x.

Copilots Jul 8 · 2-min read

GitHub Copilot adds enterprise MDM settings and OTel export

Orgs can mandate OTel telemetry destinations and push managed Copilot settings (permissions, model, plugins) via Intune, Jamf, Group Policy, or Ansible for org-wide control.

Automation Jul 8 · 1-min read

Cloudflare Workers AI toMarkdown and AI Search add GIF and BMP

toMarkdown and AI Search now ingest .gif and .bmp via object detection into a vision model, indexed automatically on next sync with no config. Stop pre-filtering these formats.

Copilots Jul 7 · 2-min read

JetBrains AI Assistant adds agent customization: hooks, MCP, skills

Agent Customization hooks, in-IDE MCP server management, and /create-skill or prompt or agent generators turn JetBrains into a configurable multi-agent IDE rather than a single-agent chat.

Agents Jul 4 · 2-min read

Windsurf and Devin surface reviewable diffs in autonomous mode

Devin Local shows autonomous edits as reviewable diffs with in-client ACU usage, making autonomous edits auditable and cost-visible. Review diffs before merge and watch the ACU meter.

OpenSource Jul 7 · 2-min read

LongCat-2.0: open-source 560B reasoning model breaks onto Product Hunt

A 560B-param open-weight reasoning model optimized for speed charted on Product Hunt. One to evaluate for a self-hosted reasoning stack, but verify weights and license before production.

Models Jul 9 · 3-min read

OpenAI ships GPT-5.6 GA: Sol, Terra, Luna tiers

All three tiers are live in the API, ChatGPT, and Codex; Terra is roughly GPT-5.5-class at about half the cost ($2.50/$15 per 1M).

Models Jul 8 · 3-min read

xAI releases Grok 4.5, "Opus-class" and cheaper

A frontier coding/agent model at $2/$6 per 1M, available day one in Cursor on every plan and in the xAI console. Not in the EU yet.

Agents Jul 9 · 3-min read

OpenAI "ChatGPT Work": an hours-long cross-app agent

Blends Codex and ChatGPT to act across apps and files and stay on a single task for hours at a time.

Voices Jul 8 · 2-min read

OpenAI GPT-Live: new listen-and-speak voice models

New listen-and-speak-at-once models power ChatGPT Voice; API access is "coming soon," so it is ChatGPT-only for now.

Copilots Jul 8 · 2-min read

Grok 4.5 lands in Cursor across every surface

Selectable today in desktop, web, iOS, CLI, and SDK, for multistep tasks at $2/$6 per 1M.

Copilots Jul 9 · 2-min read

Codex is now built into the ChatGPT desktop app

The macOS and Windows client gets inline code editing, annotations, and GitHub PR support.

Automation Jul 6 · 3-min read

Cloudflare Workers Cache: zero CPU billing on hits

One line of Wrangler config puts a tiered regional cache in front of any Worker; cache hits bill no CPU.

Models Jul 8 · 2-min read

Cloudflare adds Moondream 3.1 vision model to Workers AI

A sub-second edge VLM (query, caption, point, detect) callable via env.AI right at the edge.

Voices Jul 7 · 2-min read

AssemblyAI Universal-3.5 Pro: best-in-class async STT

Joint transcription plus diarization (30.17% cpWER vs Nova-3's 37.92%) across 18 languages.

Launch Jul 9 · 2-min read

Vercel adds one-click deploys for Lovable apps

Vibe-coded Lovable projects get git-synced production hosting on Vercel.

Agents Jul 3 · 2-min read

Replit Agent wires payments; new native Desktop App

Ask the Agent to "add payments" and it creates and connects Whop; plus a new native desktop app.

OpenSource Jul 7 · 4-min read

Hugging Face LeRobot v0.6.0: biggest robotics update yet

World-model policies, a reward-model API, six sim benchmarks, and FSDP cloud training.

Automation Jul 9 · 2-min read

Supabase adds searchable field-level encryption

Encrypt PII, health, and financial columns via CipherStash and still query them.

Copilots Jul 9 · 2-min read

GitHub Copilot generates repo overviews on any repo page

"Give me a high-level overview" produces purpose, stack, and a contributing guide (or a README) for fast codebase onboarding.

Models Jul 9 · 2-min read

fal serves sub-second Ideogram v4 with no quality loss

A 6x speedup (2.75s to 0.44s) via FP4 quant plus kernel fusion; new Base/Fast/Instant tiers for image-heavy apps.

Automation Jul 2 · 2-min read

Linear adds OAuth manifests + enterprise-managed MCP auth

One-click OAuth app setup and governed MCP access across servers in Claude, without per-server login.

Copilots Jul 9 · 1-min read

Figma Make now runs on GPT-5.6

Design-to-app builds get a stronger model for more accurate first passes.

Models Jul 8 · 2-min read

Mistral Robostral Navigate: 8B embodied navigation model

A single RGB camera plus a plain-language instruction moves a robot through a space. Access is gated ("talk to our team").

Automation Jul 9 · 2-min read

Mistral Studio adds version control for prompts and skills

Versioning, ownership, lineage, and traceability, a system of record for teams shipping controlled deployments.

Copilots Jul 1 · 2-min read

GitHub Copilot CLI adds auto model selection + credit caps

The CLI routes model choice by task and lets you set per-session AI-credit limits for safer, cheaper scripted agent runs.

Get the ones that matter, weekly.

Stay on top of what shipped in AI. Weekly update.

Unsubscribe anytime
~/subscribe $