Write CHANGELOG only when cutting a release - #848
Merged
Merged
Conversation
Agents were adding long per-PR changelog entries. Release notes are now written once, at cut time, by the release agent from the merged PR log and the code. AGENTS.md (imported by CLAUDE.md) tells every agent to leave CHANGELOG.md alone, and the release runbook says where the notes come from. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
# Conflicts: # docs/releasing.md
3 tasks done
Tanmay Singla (Tanmay182003)
approved these changes
Oct 5, 2026
This was referenced Oct 5, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
PRs stop editing
CHANGELOG.md. Release notes get written once, at release time, by the release agent: the release train'srelease-syncroutine, or whoever cuts a release by hand. It works from the PRs merged since the last tag and from the code they changed.AGENTS.md(new, imported by a one-lineCLAUDE.mdso Claude Code sessions and cloud routines load it automatically): never touchCHANGELOG.mdoutside arelease/v*orrelease-syncPR, and restore it if a PR you work on already changed it.CHANGELOG.mdpreamble: replaces Release train PR 1: release.py core (offline stamp, rc versions, CHANGELOG sync/fold, blocker gate) #643's "Add entries under[Unreleased]" with "PRs never edit this file; only the release agent writes[Unreleased]". The rule that###headings set the bump level is unchanged.docs/releasing.mdstep 1: the[Unreleased]notes are written at cut time fromgit log --first-parent <last-tag>..origin/mainand the code.Labeled
agent:needs-humanso the PR burn-down routine leaves it alone. Its new rule would otherwise revert this PR's CHANGELOG.md edit. The routine prompts and the shared procedure docs onbughunt/ledgerandarch-audit/ledgeralready carry the same rule.Follow-up for the release-train design (not changed here): DESIGN.md §3.7 step 2 still describes the routine as a gap-filler, capped at 20 bullets for commits that "have no CHANGELOG change". Under this policy it becomes the only writer, so the cap and the gap-fill framing need revisiting.
🤖 Generated with Claude Code