agentleFS
claude-worldclaude-world/cf-browser/skills/changelog-monitor/SKILL.md

changelog-monitor

Skill9 starsChanged 8 months agoMITView on GitHub

Monitor release notes and changelogs from any software project via its web page

The file

Files
SKILL.md

Changelog Monitor

Track releases, changelogs, and breaking changes from any software project's web page.

Prerequisites

CF Browser MCP server must be configured (see the CF Browser quick start).

Why This Skill?

GitHub releases, framework changelogs, and library docs are often JS-rendered pages that WebFetch can't read. This skill uses browser_json to extract structured release data from any page.

Workflow

Step 1: Determine Source

From the user's input, resolve the changelog URL:

  • GitHub project → https://github.com/{owner}/{repo}/releases
  • npm package → https://github.com/{owner}/{repo}/releases (resolve via npm registry)
  • Direct URL → use as-is (e.g., https://nextjs.org/blog)
  • "latest changes to X" → search for the project's releases page

Step 2: Extract Releases

browser_json(url, prompt="Extract the latest N releases with version, date, and key changes as JSON array")

For changelog pages (non-GitHub):

browser_json(url, prompt="Extract version entries with version number, release date, and list of changes")

Step 3: Analyze

For each release:

  • Flag breaking changes (look for "BREAKING", "migration", "deprecated")
  • Flag security fixes (look for "CVE", "security", "vulnerability")
  • Categorize changes: features, fixes, performance, deps
  • Note if any changes affect the user's project (if context available)

Step 4: Report

Output a structured summary:

## {Project} — Latest Releases

### v2.1.74 (2026-03-12)
- **NEW**: Added `autoMemoryDirectory` setting
- **FIX**: Fixed memory leak in streaming API response buffers
- Action needed: None

### v2.1.73 (2026-03-11)
- **NEW**: Added `modelOverrides` setting
- **FIX**: Fixed freezes and 100% CPU loops
- Action needed: Update recommended (stability fix)

### v2.1.72 (2026-03-10)
- **NEW**: Tool search bypasses third-party proxy gate
- Action needed: None

Parameters

  • url or project (required): GitHub repo, npm package name, or direct changelog URL
  • count (optional, default=5): Number of releases to fetch
  • baseline_version (optional): Compare against a version supplied by the user or already recorded in the current workspace. This skill does not create a persistent scheduler by itself.

Examples

"What's new in Claude Code?"
→ browser_json("https://github.com/anthropics/claude-code/releases") → structured summary

"Check for breaking changes in Next.js 15"
→ browser_json("https://nextjs.org/blog") → filter breaking changes

"Latest Hono releases"
→ browser_json("https://github.com/honojs/hono/releases") → version + changes

"Has Astro released anything since 5.18?"
→ Extract releases → filter versions > 5.18

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.