ENGINEERING

Triage merged PRs into a customer-facing changelog draft

On each merge to main, an agent decides whether the change is customer-facing; if so, it drafts a plain-language changelog entry and appends it to a pending-release doc…

CategoryEngineering
Enginepaperclip
Difficultyadvanced
Triggerevent
Steps4
Setup~25 min

How it runs

The automated pipeline, trigger to output.

  • TriggerPR merged to mainGitHubGitHub
  • LogicAgent classifies customer-facing vs internal
  • ActionDraft plain-language changelog entry
  • OutputAppend entry to pending-release docConfluenceConfluence

What it does

Whenever a PR merges to main, an agent reads the PR title, description, and labels to judge whether the change is something a customer would notice (a new feature, a visible fix, a behavior change) versus internal-only work like refactors or test changes. For customer-facing changes it rewrites the technical summary into a clear, benefit-oriented changelog line and appends it to a running pending-release document for a human to approve before the next release.

When to use it

When you maintain a public changelog and the hard part is continuously deciding what is worth telling customers and phrasing it in their language. Best for teams that merge frequently and want the draft built up incrementally rather than scrambled together at release time.

How it works

  1. 1A GitHub webhook fires when a PR is merged into main.
  2. 2An agent classifies the change as customer-facing or internal using the PR content and labels.
  3. 3A logic branch drops internal-only changes and stops.
  4. 4For customer-facing changes, the agent drafts a plain-language entry from the technical details.
  5. 5It appends the entry, tagged as unreleased, to the pending-release Confluence page for human review before publishing.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect GitHubRepos, issues, pull requests, actions.
  2. 2
    Connect ConfluenceSpaces, pages, blueprints.
  3. 3
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  4. 4
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  5. 5
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.