What changed in AI tooling this week: releases, pricing, outages, acquisitions. Every item links to its source.
HydraFusion’s research preview now draws from Claude Opus 5.5, GPT 6 Astra, GPT 6.1 Sol, and GPT 6 Luna, tailors reasoning effort per stage, and upgrades Cascade with a judge/repair step and Critique…
Pantheon opened a public beta for its MCP server so agents in Claude Code, Cursor, and Copilot can list sites and environments, pull build and runtime logs, spin up Multidevs, and trigger rebuilds or…
Claude Code 2.1.300 adds a prompt.change mod hook for non typed prompt edits (history recall, clear, undo, plugin fill), $.ui.view({ agentId }) to show an agent transcript in the main view, and OpenT…
NeuBird launched GA for its Agentic Reliability Center (ARC): one governed connection to telemetry and LLMs, shared operational memory, and human in the loop policy gates for every agent that touches…
Pluralsight launched a read only AI MCP that lets Claude Code, GitHub Copilot, Cursor, ChatGPT Enterprise, Gemini, and Copilot Studio pull its courses, labs, and skills data to build learning guides…
AISLE’s October rollout adds an MCP server that exposes open repo findings to Claude Code and Cursor so agents can fix issues without leaving the editor; merges to main auto mark them resolved, and e…
gh aw v0.91.10 cuts network access for read only agent workflows by default and records plugin installs and MCP config in session provenance. Follow up releases through v0.91.14 (Oct 14) mask OTLP cr…
Anthropic opened a public beta that exports OpenTelemetry traces from Claude Managed Agents, the Messages API, and Claude SDKs to any OTLP tool—covering agent turns, model calls, tool use, and MCP ca…
Apigee hybrid 1.17.0 turns OpenAPI 3.0 specs into MCP tools via an MCP Discovery Proxy, with the data plane in your own cluster so tools/call traffic never leaves your infrastructure. Enable with ena…
Kerno v1 lets your coding agent map and test full user flows, HTTP routes, MCP tools, and queue consumers against your real databases and services, with OWASP based security checks and tests saved to…
Chainguard opened GA for enterprise on four OAuth MCP servers— cg oci , cg apk , cg versions , and cg api —so agents in Claude Code, Cursor, and other MCP clients can verify image tags, Wolfi package…
On Oct 14, OpenAI retired GPT 5.5 from ChatGPT, ChatGPT Work, and Codex on every plan; the OpenAI API is unaffected. Codex users on ChatGPT sign in should replace gpt 5.5 with gpt 6 sol (Plus/Pro/Bus…
The gh actions lock CLI extension writes .github/workflows/actions.lock , pinning every action (transitive ones included) to a reviewed commit, so a moved or hijacked tag can't swap in new code; Depe…
Vani Code Agent (preview, code install extension AhuraSense.vani code ) reads and edits your workspace with per edit diffs, runs commands only after you approve, and offers Ask, Accept edits, or Plan…
Vercel’s Oct 13 changelog puts Google’s Gemini 4 Argon on AI Gateway as google/gemini 4 argon (1M context, thinking levels low/medium/high, provider pricing with no markup). Run vercel ai gateway set…
Chainguard opened a beta for Workspaces: Chainguard hardened microVMs where developers and agents work via chainctl develop , with working tree sync, warm snapshots, host side short lived credential…
Chainguard opened a beta waitlist for Harness: it rebuilds your app from source in an isolated, no egress microVM, lets Claude Mythos 5.1 probe the live instance for exploitable vulns, and ships only…
A new repository creation:write permission lets an app, agent, or fine grained PAT create repositories without repository administration:write (which also allows delete and transfer), and installatio…
Secret scanning now detects anthropic service api key and anthropic user api key , and push protection blocks commits containing them by default on any repo with secret scanning on, free public repos…
Now in beta, Checks lets coding agents run CI mid session in credential free, egress restricted microVMs from warm dependency snapshots, with failure logs fed back into the agent's context. Each pass…