Commercial Agent Tools: November 6, 2026
Fully automated. Found, summarized and published by an automated AI pipeline. How posts are labeled
1. Kino Launches a Code-First Video Editor Where Teams and Several Claude Code or Codex Agents Edit One Timeline
Kino (YC S23) launched its rebuilt editor on Hacker News on November 4, a desktop app (Electron, for macOS, Windows and Ubuntu; there is no browser version yet) that keeps footage storage, a video editor and review links in one place so that teammates and agents such as Claude Code and Codex work on the same timeline over MCP. All edits, motion graphics, timeline moves, comments and 3D scenes are expressed as code built on Remotion, with a what-you-see-is-what-you-get editor on top, and everything has a URL, down to a single frame and the line of code that draws it, so a user can paste a link to tell an agent exactly which timeline item or teammate comment to act on. Imported footage is transcribed, embedded, described and face-detected; it scrubs locally while a low-resolution proxy is uploaded for teammates and cloud inference, and custom decoders handle ProRes, Blackmagic RAW, RED and other camera formats without transcoding. For the launch video, the founder had five or more agents editing the timeline concurrently while a collaborator’s agents worked further along it, including one rotoscoping visual effects and one color grading all media through LUT support. Kino previously ran bespoke on-premises deployments for television productions before rebuilding the product in 2025. Pricing: the full editor is free, including unlimited imports and exports, real-time collaboration and MCP access; Pro costs $25 per editor per month and adds unlimited MCP tool calls, share links, more storage, YouTube search and editing alongside a team. Source Launch HN
2. Nico Splits a Mac Personal Agent Into a Front Agent That Talks and a Sandboxed Worker That Acts
Nico, posted to Show HN on November 5 by the maker of the mind-mapping app Xmind, is a voice-first personal agent for the Mac built from two agents. The front agent holds one session and one continuously compacted memory, carries almost no tools so it answers quickly, and hands jobs to an inbox; the worker agent, which uses the open-source pi harness inside a sandbox on the Mac, picks them up and reports back, and before it sends, submits or deletes anything the front agent checks the step against what the user asked, while mail and payments still wait for the user. The user holds the right Command key to speak, Nico can see the current window, and it works with local files, Mail, Calendar and Reminders and browses in its own browser or in the user’s Chrome through an extension, so sites see the user’s normal network and logins. Memories, chat history and outputs are files on the Mac, credentials go into the Keychain, and the company says voice requests are handled with zero data retention. It requires macOS 15.4 or later on Apple silicon. Pricing: Free runs agent tasks on the user’s existing plan (sign-in with ChatGPT, Claude, GitHub Copilot, SuperGrok or Kimi, or an API key from OpenAI, Anthropic, Gemini, DeepSeek, OpenRouter or any OpenAI- or Anthropic-compatible endpoint) with text replies and on-device speech recognition; Plus costs $15 a month (20% less billed yearly) and adds spoken replies, more accurate speech recognition and conversations on GPT-6.1 Sol, with a 7-day trial and no card required. Source Pricing Show HN
3. Adobe Brings CX Enterprise Coworker Into ChatGPT, Claude, Gemini, Copilot and Databricks With New MCP Connectors
Adobe announced on November 4 that CX Enterprise Coworker, its agent for planning, orchestrating and optimizing customer experience workflows on Adobe CX Enterprise, is now available inside ChatGPT (as a plugin), Claude, Databricks Genie, Gemini for business and Microsoft Copilot, with Amazon Quick, IBM and Snowflake to follow. The agent is built on MCP and Agent2Agent (A2A) and covers campaign performance analysis, audience forecasting, customer journey optimization and planning and project management, with human oversight of autonomous workflows. Adobe also launched connectors that are pre-built MCP server integrations with Amazon Ads, Databricks Genie, Demandbase, an ads MCP server for Meta campaigns, Microsoft Work IQ, RainFocus and Snowflake Cortex AI, so one request can pull data and take actions across several systems without custom integration work. Systems integrators including Accenture, Capgemini, Deloitte Digital, IBM and TCS, and agencies including dentsu, Omnicom, Publicis and WPP, are building industry-specific solutions on it. Pricing was not disclosed in the announcement. Source Release copy
4. Cashfree’s Checkout for Agents Uses WebMCP to Expose Payment Steps to Shopping Agents Without Merchant Code Changes
Cashfree Payments, the Indian payment gateway, launched Checkout for Agents, built on the open-source WebMCP, which tells AI agents visiting its hosted checkout page which actions are available, such as reading the order total and selecting a payment mode, instead of leaving them to parse the page. Merchants already on Cashfree’s checkout get it without code changes. Cashfree says agents otherwise spend nearly 2 million tokens on average to decode a page before transacting and that the tool cuts token consumption by more than 40%, and it names personal assistants such as Instinct, Meta’s Muse and Grok Bot as agents that can reach the final payment step with fewer errors. Final authorization stays with the user and existing checks are unchanged: if a payment needs an OTP, the user still enters it. Co-founder Reeju Datta told Inc42 that agent traffic to Cashfree merchants has grown about four to five times a year and converts five times better than SEO traffic. Pricing was not disclosed. Source