changelog-capture
Run the following command in DeepSeek Harness:
dsh plugin install ChenneyZhuang/changelog-capture
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install ChenneyZhuang/changelog-capture in your DeepSeek Harness terminal to install the plugin; the source repository is https://github.com/ChenneyZhuang/changelog-capture .
About this plugin
The changelog nobody writes costs everyone: the teammate who misses a breaking change, the user who cannot tell whether to upgrade, and you, six weeks later, reconstructing from a blurred memory what a release actually changed. changelog-capture turns that release-day scramble into a one-minute habit while the diff is still in hand, so the release note assembles itself.
Each entry states Who (which users are affected), What (one sentence of behavior), and How to upgrade (the action required, or nothing). Entries are sorted into Added, Changed, Fixed, and Breaking, with Breaking always at the top. At release time every entry is verified against the actual diff: the log claims what the diff does, any user-facing difference in the diff gets an entry, and internal refactors with no observable change stay out.
It suits small teams and solo maintainers who ship user-facing changes and do not want to spend release night reconstructing a release note from memory. It does not write marketing copy; it produces an honest, diff-checked changelog where Breaking changes are never buried.
Use Cases
- Jot a one-line entry after each change; the changelog assembles itself at release time
- Breaking changes always sit at the top, never buried among routine fixes
- Every entry is verified against the actual diff before shipping
Best For
- Small teams shipping user-facing changes who hate release-night changelog sprints
- Solo maintainers who want an honest, diff-verified change list
- Project maintainers who have been asked "what changed?" more than once
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.