Skip to content
Development
Skill

/changelog

Render release notes / a changelog from the cladding spec. Use when the user asks for release notes, a changelog, 릴리즈 노트, 변경 이력, or "what changed (since <ref>)" — run `clad changelog --json` for the deterministic shipped-changes manifest, then write the human-facing notes FROM

From plugin
cladding
1619 skills6 agents1 command1 MCP
Install
$ npx -y skills add qwerfunch/cladding --skill changelog --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/changelog

Context preview

The summary Claude sees to decide when to auto-load this skill.

Render release notes / a changelog from the cladding spec. Use when the user asks for release notes, a changelog, 릴리즈 노트, 변경 이력, or "what changed (since <ref>)" — run `clad changelog --json` for the deterministic shipped-changes manifest, then write the human-facing notes FROM

SKILL.md

changelog.SKILL.md
description: Render release notes / a changelog from the cladding spec. Use when the user asks for release notes, a changelog, 릴리즈 노트, 변경 이력, or "what changed (since <ref>)" — run `clad changelog --json` for the deterministic shipped-changes manifest, then write the human-facing notes FROM it, sourcing every claim from a feature title or acceptance sentence. For audit asks, print the --audit table verbatim. Activate only when the connected project contains spec.yaml or the user explicitly names Cladding; ignore ordinary requests in uninitialized projects.

Cladding changelog — release notes from the spec

The spec is the SSoT, so "what shipped?" is *collected*, never recalled. `clad changelog` turns `<since>..HEAD` into a deterministic manifest: feature shards classified (`added-as-done` / `flipped-to-done` / `modified-while-done` / `archived`), grouped by capability (plus an `uncategorized` bucket — itself a drift signal), the spec inventory diff, and conventional `feat:`/`fix:` commits that name no feature id (work that shipped *outside* the spec — report it honestly, don't hide it).

Protocol

1. **Collect.** Run `clad changelog --json --since <ref>` (omit `--since` to use the latest tag; if the repo has no tags the command says so — ask the user for a ref). 2. **Render EN + KO release notes from the manifest, in the CHANGELOG.md house style:**

  • Open with a `**In one line:**` abstract (what a user gains, one breath).
  • Group by the manifest's capability groups; use **user-impact verbs** ("the gate now runs

your entry point"), not implementation verbs ("refactored stage runner").

  • **No `F-…`/`AC-…` ids in prose** (Soft Shell policy) — ids belong to the audit surface only.
  • **Source every claim from the manifest** — a feature `title` or an `acceptance[]` sentence.

Never invent a change, never embellish beyond what an AC states. If the manifest is empty, say "no shipped changes since <ref>" — that line is the honest deliverable.

  • Put `unsharded_commits` under an "Also changed" section, marked as not yet spec-tracked.

3. **Audit asks** ("show verification", "어디까지 검증됐어?") → print `clad changelog --audit --since <ref>` **verbatim** — that table keeps ids and marks every verification ref resolved ✓ / unresolved ✗ (a ✗ is spec-annotation drift worth reporting). 4. **Catalog asks** ("what does this project do, in full?") → `clad changelog --catalog` prints the whole capability → feature → acceptance listing of the living spec.

clad changelog                       # markdown since the latest tag (deterministic fallback)
clad changelog --json --since v0.5.2 # the manifest you render notes from
clad changelog --audit --since v0.5.2
clad changelog --catalog
Read more
Ships withcladding

For an organization to trust AI with its code, three things must hold — trust, traceability, and stability at scale. cladding wraps your AI coding agent: your intent goes in before it writes, and the result is verified against your spec after, so those three are earned, not assumed. First L4 implementation of the Ironclad standard.

Get the whole plugin

Other skills on cladding.