release-watch

name	release-watch
description	Use when an agent needs to monitor watched software releases, classify version updates, avoid duplicate release notifications, or decide whether a watched tool should be upgraded.

Release Watch

Overview

Use this skill when the user asks whether watched software, frameworks, tools, or repositories have new versions and whether those versions are worth upgrading.

Keep this workflow separate from public trend discovery. GitHub releases are upgrade-monitoring signals, not Trend Radar inputs.

Workflow Selection

• For watched software version updates, use this skill.
• For public ecosystem trend cards, use public-trend-radar.
• For ordinary subscribed RSS daily news, use rss-ai-digest.
• For source quality and registry maintenance, use rss-source-curator.
• For deep research synthesis from evidence, use subscription-research-agent.

Runtime Command

Examples use subscription-research as a compatibility command. Before running commands, resolve the runtime command:

• When DIGESTPILOT_BACKEND_URL is available, product actions such as release watch refresh or WebUI-backed upgrade monitoring may use the Backend task API first, then read the resulting workspace output.
• If the Backend API is unavailable, keep using the local runtime command fallback below. Do not require WebUI, cloud, or a hosted service in generic Agent environments.
• Use DIGESTPILOT_RUNTIME_CMD when the environment provides a custom runtime command.
• Use digestpilot-runtime when it is available on PATH.
• Use subscription-research when only the legacy compatibility command is available on PATH.
• From a repository checkout without a linked command, replace subscription-research with node packages/research-cli/dist/src/cli.js.
• Run node scripts/doctor.mjs from the repository root to diagnose local runtime setup.
• If no runtime command is available but this repository or plugin package includes runtime source, run node scripts/bootstrap-runtime.mjs, then retry the command.

Core Command

Prepare a watchlist:

{
  "repos": [
    {
      "id": "vercel/ai",
      "importance": "high",
      "current_version": "v5.1.0",
      "used_capabilities": ["streaming", "tool-calling"]
    }
  ]
}

Fetch live GitHub releases and render upgrade cards:

subscription-research release watch \
  --watchlist release-watchlist.json \
  --state release-seen.json \
  --mark-seen \
  --format markdown \
  --output release-watch.md

By default, the runtime fetches recent GitHub releases for each watched repository. Set GITHUB_TOKEN when higher GitHub API limits are needed, or pass --github-token-env NAME to read a different token variable.

For offline tests or exported data, pass prepared GitHub release JSON:

subscription-research release watch \
  --watchlist release-watchlist.json \
  --github-releases github-releases.json \
  --state release-seen.json \
  --format markdown

Prepared release JSON uses this shape:

{
  "repositories": [
    {
      "repo": "vercel/ai",
      "releases": [
        {
          "tag_name": "v5.1.1",
          "name": "v5.1.1",
          "html_url": "https://github.com/vercel/ai/releases/tag/v5.1.1",
          "published_at": "2026-06-12T00:00:00Z",
          "body": "Release notes..."
        }
      ]
    }
  ]
}

Output Guidance

Release Watch output should be an upgrade queue, not a trend list.

Use these priorities:

• upgrade-now: security fixes or urgent fixes that affect watched capabilities.
• review: breaking changes, major updates, or updates touching watched capabilities.
• watch: prereleases or unclear impact.
• ignore: maintenance-only or documentation-only releases.

Markdown output should focus on actionable releases. JSON preserves all cards, including ignored releases, for automation.

When current_version is present in the watchlist, Release Watch adds version_delta, upgrade_action, and current-version context. Use this to distinguish patch upgrades from minor upgrades and major migration reviews.

Boundaries

Do not modify user dependencies, create upgrade PRs, or publish notifications automatically. Produce reviewable upgrade cards and let the user decide the next action.