AI Product Updates: New Features and Releases

Odessia's concierge agent books trips and reservations over text, keeps watching fares and tables after you ask, and claims accredited travel agent status for agent rates.

Build-A-Chip-Bench tests whether AI agents can autonomously design chips; the best agent scored 16.6%, with about 40% of tasks passing functional verification.

The Claude Code desktop app now starts sessions faster, makes it easier to split work across several, and shows what each session is doing.

Claude Code 2.1.300 ships 88 CLI changes and one system prompt change, adding virtualText to keep draft text out of model input, $.ui.view to review agent transcripts, and a prompt.change hook.

YC-backed Instinct says it is the first frontier agent harness built to give individual traders the organizational capabilities of an investment firm.

Scope records what each AI agent on your site was asked to do, which pages it read, where it got stuck, the fix, and whether it converted after the fix shipped.

Scope identifies which AI agent visited, what it came to do, the step where it stopped, and whether it converted.

Journeys detects when AI agents visit a product, follows their paths and purchases or installs, pinpoints blockers, and lets you publish prepared fixes with one click.

Odessia launched today as an AI agent that is also an accredited travel agent, booking flights, hotels, tables and tickets by text, using points, and rebooking automatically when prices drop.

Odessia is a travel concierge you text that books flights, tables and tickets, uses Amex or Capital One points, and keeps watching for sold-out seats and fare drops.

A roundup of eight updates: GPT-5.5 leaving ChatGPT, Gemini 4 Argon pricing, a combined Grok and Cursor subscription, Claude in Google Docs, and more.

LangChain launched a public beta that exports traces from Claude Managed Agents, the Messages API and the Claude SDKs to LangSmith through native OTel support.