AI Product Updates: New Features and Releases

Recording is live in dMeetApp, and the dTelecom x402 gateway has processed over $104K in volume across 49K+ transactions, letting agents call voice, STT and TTS with no API keys or accounts.

Binance released three AI products, with Pro letting agents connect directly to trading accounts, generate strategies from one sentence, use a visual strategy editor, and access all in-platform data.

Fulcra moves context out of individual apps into a store the user owns, shared by Claude, ChatGPT and Grok over MCP, free up to 5 GB with Pro at $14.99/month.

The project shipped an AI agent, an API, a GitHub App, a dashboard and an openclaw plugin in September, with 11 releases and 4 new partners signed.

AI workspace KooKo launched with a single agent that handles slides, spreadsheets, docs and design and remembers brand settings, so assets only need to be set up once.

Open-source Watchtower launched as a self-hosted alternative to Sentry and AppSignal, with Slack, Linear and email alerts, an MCP server for coding agents, and a single Go binary plus Postgres.

Equifold launched Agents on Robinhood Chain, letting tokens create AI agents with a mission, a choice of 254 models and a real browser for public research, with compute funded by the token's trading fees.

Lexington Themes released the Navy template with pricing, status, security, integrations, changelog and help center pages, available for Astro, Sanity, EmDash CMS and MCP.

JONI combines several leading AI models to write copy, build a site, set up domain and hosting and publish it, asking for user approval before key actions, with site builds costing about $100.

Vangrid is live, letting agents order physical-world data capture through USDC bounties on Base and Arc, with humans doing the capture and agents verifying the work before onchain settlement.

Prediction market Limitless launched MCP, letting ChatGPT, Claude and Grok filter markets, read probabilities, analyze trade history and prepare orders, alongside a Profile and leaderboard rebuild.

Uber released ADR, its internal AI agent security system that discovers unapproved AI tools, observes agent behavior and detects threats, along with a 300+ task ADR-Bench; it already runs in production.