What AI Update Did You Miss Today?

GitHub Copilot: Request Code Reviews Through the REST and GraphQL APIs

GitHub Copilot code review can now be requested through the **REST and GraphQL APIs**, with an optional **review effort level** set per request. This lets teams start Copilot reviews from their own scripts, workflows, and internal tools. The feature is generally available to Copilot Pro, Pro+, Max, Business, and Enterprise plans, and the post also restates that **Balanced** is the default effort level.

GitHub Copilot

Replit Adds GPT-6.1 Sol and Claude Sonnet 5.5 to Agent

Replit added two new model options to Agent: **GPT-6.1 Sol**, selectable in **Max mode**, and **Claude Sonnet 5.5**, selectable in **Power mode**. The release gives developers a newer OpenAI and Anthropic model to pick from when building with Agent. Replit also published a blog post explaining its philosophy on model selection.

Replit

Replit Adds Jev Decision Model to AI Integrations

Replit added **Jev**, a decision model available as `jev-latest` through OpenRouter inside **Replit AI Integrations**. It lets apps built on Replit **classify content, route requests and score leads** with fast, structured AI decisions. Developers do not need to manage their own API keys, and they can ask Agent to turn it on.

Replit

Claude Code Closes rm, BASHPID and Fail-Open Hook Gaps in Permission Checks

Claude Code 2.1.288 patches four gaps where the agent could act outside what the user authorized. A dangerous `rm` inside a `bash -c` script could run unprompted in bypassPermissions mode or under a shell allow rule, a `BASHPID` arithmetic assignment slipped past the prompt, hooks that failed to match were skipped, and sandbox credential file entries on git config files were ignored under one setting. Updating closes all four.

Claude Code

Claude Code Lifts Background Command Time Limit Outside Unattended Sessions

Claude Code 2.1.288 narrows the background command time limit introduced days earlier so that it applies only to unattended sessions such as `-p`, the Agent SDK, CI and cloud runs. Interactive sessions in the terminal, the desktop app and VS Code now have no limit. The change restores long-running dev servers and watchers for people working at the keyboard while keeping a safety net for automation.

Claude Code

Claude Code Lets You Recover a Prompt Cleared With Ctrl+C

Claude Code 2.1.288 adds recovery for a prompt that was cleared with Ctrl+C. Pressing Up on the empty prompt now brings the draft back, including pasted text and images. The change removes a long-standing source of lost work, especially when composing long prompts with screenshots.

Claude Code

Claude Code: Mods Let Plugins Rewrite Prompts, UI and Built-in Features

Anthropic launched Claude Mods on October 1, 2026, letting plugins ship small TypeScript functions that hook into Claude Code events. A mod can rewrite prompts, block or retry tool calls, answer permission requests, redact tool output and replace interface elements. Some built-in features, such as /diff, now ship as mods, and a first built-in mod called You should know runs a side agent that flags things the user or Claude might miss.

Claude Code

Claude Code: Shell Write Safeguards Tightened for rm, Symlinks and Credential Files

Claude Code 2.1.287 closes several gaps where shell commands could write outside what the user approved. A dangerous rm no longer loses its always-ask safeguard when it also redirects output to a home or wildcard path, writes through repo-committed symlinks now wait for a person, and broad Bash allow rules no longer let shell writes reach credential files that Claude's file tools refuse.

Claude Code

GitHub Copilot: Computer Use for Desktop Apps in the CLI and the Copilot App

GitHub introduced **computer use** in public preview for **GitHub Copilot CLI** and the **GitHub Copilot app** on macOS and Windows. Copilot can now read accessible app content and visual context, click controls, type, press keys, scroll, and drag across desktop applications, including **legacy and GUI-only software** with no API, command line, or MCP integration. Copilot asks for approval before controlling an app, and organization-managed settings can disable the feature.

GitHub Copilot

GitHub Copilot: Dynamic Workflows in Copilot CLI and the Copilot App

GitHub introduced **dynamic workflows** in public preview for **Copilot CLI**, the **GitHub Copilot app**, and the **GitHub Copilot SDK**. A dynamic workflow is a program that defines a task in code, combining automated steps with one or more agents that run sequentially, in parallel, or both. This gives complex multi-agent work **repeatable stages, checkpoints, and observability**, and it is available on all Copilot plans.

GitHub Copilot

Lovable: Live Voice Conversations With GPT Live

Lovable now lets builders add **live voice conversations** to their apps using **GPT Live**. Users can talk to an app, interrupt it mid-answer, and get real-time spoken responses, which suits conversational tutors and spoken assistants. The feature requires projects built on **TanStack Start** and consumes credits.

Lovable

Lovable: EU Inference Keeps AI Processing in Europe

Lovable added an **EU inference** mode that lets Enterprise workspace admins and owners route AI model requests through **European Union providers**. The setting lives in Privacy & security and applies both while building and when published apps run. Some models and features are unavailable in the EU, and credit consumption is higher.

Lovable

Base44 Workspace Members Can Auto-Join Your App

Base44 added collaboration settings that let builders choose between **Workspace members** and **Invited only** access for an app. With **Workspace members**, teammates in the workspace get access automatically.

Base44

Base44 Lets You Control Who Can Embed Your App

Base44 added an **embedding control** to every app's security settings, letting builders choose whether **anyone**, **only listed sites**, or **no one** can embed the app in an iframe. On Enterprise plans, administrators can set a single **workspace-wide embedding policy** in Governance, and individual apps can only tighten it, never loosen it.

Base44

GitHub Copilot: HydraFusion Multi-Model Orchestration in VS Code and the Copilot App

GitHub expanded **HydraFusion**, its research preview for **multi-model orchestration**, from Copilot CLI to **Visual Studio Code** and the **GitHub Copilot app**. Instead of acting as a single model, HydraFusion picks a **Single, Cascade, or Critique workflow** for each task and coordinates models from different providers within one turn. The expansion adds clearer visibility into each workflow step and more frequent progress updates during long tasks. It is available to Copilot Pro, Pro+, Business, and Enterprise subscribers.

GitHub Copilot