Add to Chrome

FOR DEVELOPERS · AI RELEASE MONITORING

Monitor AI releases.
Know what changed.

Keep a watch on AI model announcements, SDK changelogs, and MCP project releases. Review updates from chosen source pages in a history you can revisit.

Add to Chrome See the setup steps ↓

Start from a source with dated updates

A team building with AI agents may depend on a model API, a client SDK, and several tools. Pick the changelog for one dependency you already use. The Claude Platform release notes, for example, collect updates to the API, SDKs, and console.

For an open-source dependency or MCP server, open its repository's Releases page. GitHub releases organize release notes around versions. A specific Hugging Face model card can also be useful when you want to follow changes to a model's documented uses or limitations.

Look beyond the model name

  • Version and date: distinguish a new release from an edit to an older announcement.
  • Availability: note whether the source describes a preview, general release, limited rollout, or deprecation.
  • Migration work: look for renamed parameters, changed defaults, SDK requirements, or retirement dates.
  • Scope: keep the provider, product, and affected feature together so a console update is not mistaken for an API change.

This creates a useful review list for your own project. A release announcement alone cannot show whether a model or tool will work better on your tasks.

Create a changelog watch

  1. Install PageLens and open the source page in Chrome. Choose a stable changelog URL rather than a search results page.
  2. Open the extension, enter an instruction, select browser execution, and choose an interval. Daily checks are a starting point for a release review routine.
  3. Grant access to the site. Check the first saved result for the latest visible entry, its date, and its version.
  4. Review later changes against the baseline. Add a separate watch for another provider or repository when you need it.

To use AI summaries for a browser watch, configure and test your own model connection and enable analysis. Basic text-change checks work without it; this instruction does not filter the underlying change notifications.

Turn an update into a review decision

Use the saved evidence to decide whether to read a migration guide, schedule an evaluation, or update a dependency. PageLens monitors pages you choose; it does not discover every new AI model, call a repository API, test MCP compatibility, or install releases.

Pagination and collapsed entries can hide content from a check. Edits elsewhere on the page may also trigger a notification. Browser execution requires Chrome running and your computer awake. Cloud execution is available after sign-in when you need scheduled checks to continue after Chrome closes.

Instruction copied