agentsclimarketplace
Sign in

Utility pm changelog curator

Skill product-on-purpose/pm-skills/skills/utility-pm-changelog-curator

Draft CHANGELOG entries from git log via the pm-changelog-curator sub-agent, applying the repo hygiene rules (describe what changed, public paths only, no attribution trailers). Returns a layered draft with a status summary for maintainer review; refuses a dirty working tree unless --committed-only is passed. Use when banking unreleased changes or preparing a release.From its SKILL.md

Install
npx -y skills add product-on-purpose/pm-skills --skill utility-pm-changelog-curator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its file declares

Copied from the file, not written here

The file declares its own license as Apache-2.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

4.2 KB, 867 tokens by cl100k_base, as published. Nobody here has run it

<!-- PM-Skills | https://github.com/product-on-purpose/pm-skills | Apache 2.0 -->

PM Changelog Curator (Dispatch Skill)

Cross-client dispatch wrapper for the pm-changelog-curator sub-agent. Detects runtime; dispatches to the native sub-agent on Claude Code; reads agents/pm-changelog-curator.md and executes inline on non-Claude clients.

When to Use

  • You are preparing a release and want a CHANGELOG draft from git log between two tags
  • You are running on a non-Claude AI client without native pm-changelog-curator sub-agent support
  • You want a draft that respects pm-skills CHANGELOG hygiene rules (no internal-notes references, no em-dashes, public paths only) without manually applying them

When NOT to Use

  • You want to commit CHANGELOG entries directly (this skill produces a draft for review, not a commit)
  • You want to review a PM artifact -> use utility-pm-critic instead
  • You want a repo-wide governance audit -> use utility-pm-skill-auditor
  • You want a full guided release flow -> use utility-pm-release-conductor (which chains to this skill at gate G2)

Instructions

Runtime detection step. Determine which AI client is invoking this skill.

If you are running in Claude Code with the pm-skills plugin installed

Invoke @agent-pm-skills:pm-changelog-curator with the user's arguments. Pass --since-tag, --target-version, --committed-only from $ARGUMENTS if present. Return the sub-agent's draft to the user.

If you are running in any other AI client

Codex CLI, Cursor, Windsurf, Copilot, Gemini CLI, or any other client without native pm-skills plugin sub-agent support:

  1. Read the canonical sub-agent definition at agents/pm-changelog-curator.md
  2. Execute the system prompt body in that file as your operating instructions
  3. Run the 8-step drafting flow documented in the sub-agent definition (establish range -> read hygiene rules -> check working tree -> enumerate commits -> classify -> group -> rewrite -> determine target version -> emit draft)
  4. Apply --since-tag, --target-version, and --committed-only arguments from $ARGUMENTS
  5. Return the layered output per master plan D26 (full CHANGELOG draft + Status Summary + Status YAML)

Cross-Client Notes

The dispatch skill requires the AI client to:

  1. Execute Bash for git log, git describe, and git status invocations
  2. Read CLAUDE.md hygiene rules and existing CHANGELOG.md format reference
  3. Read the canonical sub-agent definition file
  4. Treat the agent body as operating instructions

See Sub-Agent Compatibility Matrix for the canonical cross-client status. Summary for this skill as of v2.16.0: PRODUCTION on Claude Code + Codex CLI (Codex CLI successfully ran git log + read CLAUDE.md hygiene rules + drafted clean CHANGELOG entries + correct dirty-tree refusal); EXPERIMENTAL on Cursor / Windsurf / Copilot CLI / Gemini CLI. If a client is unreliable, fall back to manual git log + manual drafting using CLAUDE.md as the hygiene reference.

Reference Files

What ships with it: 4 files

14.5 KB alongside SKILL.md

evals/

references/

Gives 0 of the 12 instructions most readme changelog skills give in 867 tokens

Counted across 446 of the 460 authors here whose files we hold, read 2026-09-06

  • Follow Keep a Changelog formatin 24 of 446
  • Collect commits since the last git tagin 15 of 446, across 13 files
  • Omit empty sectionsin 14 of 446
  • Put breaking changes first with migration stepsin 14 of 446
  • Include migration guidance for breaking changesin 11 of 446, across 10 files
  • Categorize commits by conventional commit prefixin 11 of 446
  • Mark breaking changes prominentlyin 10 of 446
  • Prepend the new entry to CHANGELOG.mdin 9 of 446
  • Highlight breaking changes with migration notesin 8 of 446, across 7 files
  • Classify changes into Keep a Changelog categoriesin 8 of 446, across 7 files
  • Group related commits into single entriesin 8 of 446
  • Write the changelog from commitsin 8 of 446

Said here and by no other author read

  • Detect the invoking AI client runtime
  • Invoke the pm-changelog-curator sub-agent on Claude Code
  • Read the canonical sub-agent definition file
  • Execute the agent body as operating instructions
  • Run the 8-step drafting flow
  • Apply since-tag, target-version, and committed-only arguments

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.