han-release

Pre-requisites

• gh CLI: !which gh 2>/dev/null || echo "not installed"
• jq: !which jq 2>/dev/null || echo "not installed"
• claude CLI: !which claude 2>/dev/null || echo "not installed"
• git repo: !git rev-parse --is-inside-work-tree 2>/dev/null || echo NO

If gh, jq, or claude reads not installed, or this is not a git repo: tell the operator which prerequisite is missing and that it must be installed/configured before /han-release can run, then immediately stop. The skill cannot proceed without all four.

The claude CLI is what creates the per-plugin tags in Step 10. Every invocation of it in this skill goes through the shell's command builtin (command claude ...), never a bare claude, because an operator's shell commonly wraps claude in a function or alias that blocks waiting for terminal input. The which probe above resolves the same way a bare call would, so it reports the wrapper's presence rather than the executable's; treat a non-empty result as "the tool is reachable" and let Step 10's first invocation be what proves it runs.

Project Context

• repo: !gh repo view --json nameWithOwner -q .nameWithOwner 2>/dev/null || git config --get remote.origin.url
• current branch: !git branch --show-current 2>/dev/null || echo unknown
• default branch: !git symbolic-ref --short refs/remotes/origin/HEAD 2>/dev/null | sed 's#^origin/##' || echo unknown
• working tree: !git status --porcelain 2>/dev/null || echo NO
• parent plugin name: !jq -r .name .claude-plugin/marketplace.json 2>/dev/null
• plugins (name source version): !jq -r '.plugins[] | "\(.name)\t\(.source)\t\(.version)"' .claude-plugin/marketplace.json 2>/dev/null
• latest parent tag: !git fetch --tags --quiet >/dev/null 2>&1; git tag -l 'han--v*' --sort=-v:refname | head -n1
• latest suite tag: !git tag -l 'v*.*.*' --sort=-v:refname | head -n1
• changelog head: !grep -m1 '^## v' CHANGELOG.md 2>/dev/null

The two tag probes are separate on purpose. A single pattern covering both namings returns the wrong tag: version sorting compares the whole refname, so v4.6.0 sorts ahead of han--v5.0.0 and the old suite tag wins on every release after the transition. The fetch runs on the first probe only; both read the same refreshed tag list.

latest parent tag carries the literal han--v* pattern, because a context-injection command is a fixed string and cannot interpolate parent plugin name. If parent plugin name is not han, redo the lookup in Step 2 with the actual name before using either value. Skipping that turns a renamed parent into an empty probe, which Step 2 would read as a first release and silently expand the changelog to the whole repository history.

Vocabulary used throughout this skill

• parent — the meta-plugin whose name equals the marketplace name (parent plugin name above, normally han). It has no skills or agents of its own; it exists to install the children via dependencies. The parent's own per-plugin tag is the one the GitHub release attaches to, so the release tag is {parent plugin name}--v{parent target}.

• children — every other entry in marketplace.json.plugins[] (han-core, han-github, han-reporting, and any future han-* plugin). Each child has its own version line, bumped independently of the others.

• baseline of a plugin — its version at prev (the latest release tag). For the parent this is prev#. For a child it is the version recorded in that child's plugin.json at prev; if the child did not exist at prev, it is a new plugin (see Step 3).

• current of a plugin — the version in its working-tree plugin.json.

• target of a plugin — the version being released for it. The release tag is {parent plugin name}--v{parent target}.

• tag name of a plugin — {name}--v{target}, for example han--v5.0.0 and han-core--v3.0.0. Every plugin gets one, and the parent's is what the GitHub release attaches to.

• prev is the previous release's tag, resolved in Step 2 from the two tag probes. On the first release it is empty.

• prev# is the part of prev after its last v. That rule is correct under both namings, which matters because prev can be either shape:

  v4.6.0        -> prev# is 4.6.0
  han--v5.0.0   -> prev# is 5.0.0
  

  Do not read prev# as "the number without the leading v". That older rule returns the whole tag string for a per-plugin tag, and prev# is the parent's baseline (see Step 3a), so a wrong value here silently corrupts the entire version plan rather than failing.

• Each plugin's source directory comes from the source field in marketplace.json (for example ./han-core), so its plugin.json is {source}/.claude-plugin/plugin.json. Use {source} verbatim in every git command: the ./-prefixed form works both after a {ref}: colon (git show {prev}:{source}/...) and as a pathspec (git diff ... -- {source}/). Do not strip the leading ./.

Step 1: Parse the invocation and check release safety

1. Parse $ARGUMENTS for two independent flags, then treat the remaining free text as optional release context that informs the changelog narrative:

   • pause_before_publish — true if the argument contains "pause", "review", or "confirm before publish" (case-insensitive). Default false.
   • draft_release — true if the argument contains "draft". Default false.
   • The leftover text (anything that is not those flag phrases) is $release_context, passed into the narrative dispatch in Step 5. May be empty.

2. Working tree must be clean. If working tree from Project Context is non-empty, there are uncommitted or untracked changes. Stop and tell the operator to commit or stash them first. Releasing an unknown working state is unsafe and a pushed tag is hard to reverse. This is a hard stop, not a pause gate.

3. Branch note (non-blocking). If current branch is not the default branch, do not stop — note in the Step 7 summary that the release is being cut from current branch and the tag will point at that branch's HEAD. The operator chose autonomous; surface the fact, do not block.

Step 2: Determine previous version, commit range, and PR list

1. Resolve prev from the two tag probes, in this order:

   1. latest parent tag when it is non-empty. This is the normal case from the second per-plugin release onward.
   2. Otherwise latest suite tag. This is the transition release, the one release where the previous tag still carries the old vX.Y.Z naming. Once a parent tag exists on the remote this arm is unreachable and can be deleted.
   3. Otherwise empty: this is the first release. There is no previous tag, the commit range is the full history, all compare links are omitted, and every plugin is treated as new (Step 3).

   If parent plugin name is not han, the latest parent tag probe used the wrong literal. Re-run the lookup yourself before applying the precedence above: git tag -l '{parent plugin name}--v*' --sort=-v:refname | head -n1.

   prev feeds two different things: the commit range below, and the parent's baseline version in Step 3a via prev#. Parse prev# with the after-the-last-v rule in the vocabulary block, not by stripping a leading v.

2. Commit range. With a previous tag: ${prev}..HEAD. First release: the full history (HEAD with no range base).

3. Nothing to release check. Run git log {range} --oneline. If it is empty, there are no commits since prev. Stop and tell the operator there is nothing to release.

4. Collect merged PRs in the range. Extract PR numbers from both squash subjects and merge commits:

   git log {range} --pretty=%s%x00%b | grep -oE '#[0-9]+' | tr -d '#' | sort -un
   

   For each number N, run gh pr view N --json number,title,author,url,mergedAt,state. Keep only entries where state is MERGED. Sort the survivors by mergedAt ascending (newest merge last). This is $pr_list. The PR list is repo-wide and appears once per release; it is not split per plugin. Build the PR lines and the changelog bullets per references/release-notes-format.md and references/changelog-rules.md.

5. No-PR fallback. If $pr_list is empty (local-only or squash history with no PR refs), record the notable commit subjects from git log {range} --oneline instead, and use the commits form documented in both reference files.

6. Collect closed issues and their attribution. For each merged PR N in $pr_list, find the issues that PR closed and credit everyone involved. This relates each closed issue to the fix that resolved it.

   • Find the closed issues for the PR. Take the issue numbers from gh pr view N --json closingIssuesReferences --jq '[.closingIssuesReferences[]?.number]' (the GitHub-tracked closing links). As a fallback for older PRs that linked via text, also scan the PR body and commit messages for GitHub closing keywords: gh pr view N --json body,commits --jq '[.body, (.commits[].messageBody)] | join("\n")' and extract #<num> that follow close, closes, closed, fix, fixes, fixed, resolve, resolves, or resolved (case-insensitive). Union the two sets, dedupe.

   • Confirm each is a closed issue. For each candidate number I, run gh issue view I --json number,title,author,state,comments (suppress stderr; redirect 2>/dev/null). Skip the number if the command fails (it is a PR number, not an issue, or does not exist).

   • Gather attribution per issue. Record:

     • opener — .author.login, unless .author.is_bot is true.
     • issue contributors (people who contributed meaningfully) — the people who left a substantive comment on the issue. A reaction is not a comment (a 👍 or other emoji reaction never appears in .comments[]), so reaction-only participants are already excluded. Drive-by comments do not count either. Pull each comment with its author and body (gh issue view I --json comments --jq '.comments[] | select(.author.is_bot|not) | {login: .author.login, body: .body}', stderr suppressed), and treat a comment as a drive-by when its trimmed body is emoji-only, or is a brief acknowledgment or status ping (for example +1, same, me too, bump, following, thanks, any update(s)?), or is shorter than roughly 15 words and adds no detail. A person qualifies only when at least one of their comments is substantive (not a drive-by). Remove the opener and the PR workers so each person is credited once. May be empty.
     • PR workers — for the closing PR N, the union of the PR author, the review authors, and the commit authors: gh pr view N --json author,reviews,commits --jq '[.author.login] + [.reviews[]?.author.login] + [.commits[].authors[].login] | unique'. Drop bot accounts (is_bot where available, plus the web-flow, github-actions, and dependabot logins).

   • Build $issue_list. One entry per closed issue: its number, title, opener, contributors, the closing PR number(s), and the merged PR workers. If the same issue is closed by more than one PR in the range, record every closing PR and merge their worker sets. Build the changelog bullets and release-body lines per references/changelog-rules.md and references/release-notes-format.md. If no closed issues are found, $issue_list is empty and the issues subsection/section is omitted everywhere.

Step 3: Build the per-plugin version plan

Enumerate the plugins from plugins in Project Context (one parent, plus each child). For every plugin, determine baseline, whether it changed in {range}, and its target. Classify changes against docs/semantic-versioning.md. The governing rules:

• The parent always bumps on every release. Even when only one child changed, the parent gets a version bump, because every release is a release of the suite.
• A child bumps only when its own directory changed in {range}. A child with no changes keeps its version.
• A brand-new plugin is not bumped by the release that introduces it. Its plugin.json version is its established baseline. Record the introduction in the changelog, but do not increment. This is the general rule for every future han-* extension, not a one-time exception for the current children.

3a. Classify each plugin

For each plugin, read current from {source}/.claude-plugin/plugin.json and compute baseline:

• Child, did not exist at prev (git cat-file -e {prev}:{source}/.claude-plugin/plugin.json fails, or this is the first release): new plugin. baseline = current, target = current, no bump, mark it new. Skip the rest of the classification for this plugin.
• Child, existed at prev: baseline = git show {prev}:{source}/.claude-plugin/plugin.json | jq -r .version.
• Parent: baseline = prev# (the parent's version is what the tag tracks, regardless of any directory move). On the first release baseline is empty and the parent is treated like a new plugin set to its current value.

Determine whether the plugin changed in {range}:

• Child: changed when git diff --name-only {prev}..HEAD -- {source}/ is non-empty.
• Parent: always treated as changed (it always bumps). Its change level is computed in 3b from the whole release, not just {parent source}/.

3b. Compute each changed plugin's bump level and target

For a changed child, classify the highest-priority change inside {source}/:

• major — a skill directory under {source}/skills/ was removed or renamed (renaming breaks /skill-name), an agent under {source}/agents/ was removed or renamed, or a commit indicates a breaking behavior change (! in the type, BREAKING CHANGE, a review skill that now auto-posts, and so on). Inspect git diff --name-status {range} -- {source}/ for D/R on SKILL.md or agent paths, and scan commit subjects scoped to that plugin.
• minor — a new skill, a new agent, a new references/ file, or a new optional capability was added inside {source}/, with no major change present. Inspect the same diff for added SKILL.md / agent files.
• patch — only typo, permission, edge-case, or context-injection fixes inside {source}/.

For the parent, the bump level is the maximum across the whole release:

• a child was removed from the suite → major (breaking for anyone who installed the meta-plugin).
• any changed child's level is major → major.
• a new child plugin was introduced, or any changed child's level is minor → at least minor (a new or expanded capability reaches suite installers).
• otherwise (only child patches, or only repo-level/{parent source}/ doc and config fixes) → patch.

Take the highest of those. Repo-root changes that do not live inside any plugin directory (for example docs/, README.md, CONTRIBUTING.md) are suite-level: they count toward the parent's level (normally patch) and never bump a child.

Compute proposed from each plugin's baseline: major → (x+1).0.0, minor → x.(y+1).0, patch → x.y.(z+1).

3c. Decide each plugin's target (ahead path vs. compute path)

For each changed plugin, compare current to baseline:

highest=$(printf '%s\n%s\n' "{baseline}" "{current}" | sort -V | tail -n1)

• current strictly ahead of baseline (highest == current and current != baseline): the version was already bumped during development. target = current. No confirmation for this plugin. Still compute the expected proposed, and if current is a lower level of bump than the changes warrant, add one non-blocking advisory line to the Step 7 summary.
• current equal to or behind baseline: the one-bump-per-branch bump has not been applied for this plugin. target = proposed; this plugin needs confirmation.

3d. Confirm the plan (conditional gate)

target = parent target drives the tag {parent plugin name}--v{parent target}.

This gate is conditional and often does not fire. It is not the gate that approves the tags; that one is Step 9 and it always fires.

• If no plugin needs confirmation (every changed plugin was already ahead, plus the new plugins): the plan is fully determined. Do not prompt. Record it for the Step 7 summary and continue.
• If one or more plugins need confirmation: present the whole plan in one AskUserQuestion (header: "Release versions"). State prev, and for every plugin a line of the form {name}: {baseline} → {proposed} ({level}; {evidence}), marking new plugins as new at {current} (no bump), ahead plugins as already at {current}, and unchanged children as unchanged at {current}. Name the specific skills/agents and commits that drove each level. Options: accept the proposed plan (recommended, first); adjust the parent level; adjust a child level; enter explicit versions. Apply the operator's answer to the affected plugins. Record the final plan as the version decision in the Step 7 summary.

Step 3.5: Validate every plugin before anything is written

Run this before Step 4 writes anything. The working tree is still clean here, and that is the point: a failure at this step leaves the repository exactly as the run found it, with nothing written and nothing pushed.

Both checks cover every plugin in plugins, not only the ones being bumped. Version drift in an untouched plugin is the likeliest way this fails, because Step 4 only writes plugins whose target differs from current and so never corrects a plugin nobody changed.

1. Version agreement. For each plugin, compare its plugin.json version against its marketplace.json entry selected by name:

   jq -r .version {source}/.claude-plugin/plugin.json
   jq -r --arg n '{name}' '.plugins[] | select(.name == $n) | .version' .claude-plugin/marketplace.json
   

   Any disagreement stops the run. Name the plugin and both versions. Do not attempt a repair: the correct value depends on which file is wrong, and that is the operator's call.

2. Tree cleanliness. git status --porcelain must be empty. Step 1.2 already required this, so a non-empty result here means something wrote a file since the run started. Stop and say so.

Do not implement either check by invoking command claude plugin tag --dry-run. That command refuses on a dirty plugin folder, and after Step 4 every bumped plugin's folder is dirty by construction, so a dry-run sweep placed after the writes refuses exactly the plugins the release just bumped and cannot tell that refusal apart from real drift. The two reads above answer the same question and work in either state.

Step 4: Apply the versions

For every plugin whose target differs from its current (the compute-path plugins from Step 3c, and any plugin the operator edited at 3d), set both files so they read target:

1. Set {source}/.claude-plugin/plugin.json version to that plugin's target (Edit).
2. Sync that plugin's marketplace.json entry: set the version of the plugins[] element whose name equals the plugin name (Edit). Select by name, not by index.

Skip any plugin whose target == current (ahead-path or new plugins — their files are already correct). When the entire plan is ahead-path/new (no version differs from current), this step is a no-op; note it and continue.

Re-run the version-agreement check from Step 3.5 over every plugin this step wrote. Each plugin's two files are edited separately, so this step is the one place in the run that can create a half-applied bump. Skip the cleanliness check here: the tree is dirty by design now. A disagreement stops the run, and nothing has been committed yet, so the recovery is discarding the working-tree edits.

Step 5: Update CHANGELOG.md

Follow references/changelog-rules.md exactly. From v3.0.0 onward, each release section is a parent ## v{parent target} heading with one ### {plugin} v{version} sub-heading per plugin that changed (the parent always appears; new and changed children appear; unchanged children are omitted), plus the release-level bookkeeping subsections. Every @mention in the changelog (narrative, PR bullets, issue bullets) is a markdown link to the person's GitHub profile: [@{login}](https://github.com/{login}), never flat text.

1. Does ## v{parent target} already exist in CHANGELOG.md? Search for the literal heading.

2. It exists — augment. Leave every existing line of the narrative untouched. Put the generated bookkeeping subsections as the last ### subsections of the ## v{parent target} section, before the next ## v heading, in this order: ### Issues closed in this release (only when $issue_list is non-empty), then ### Pull requests in this release (or the commits form from the fallback). Build the issue bullets from $issue_list and the PR bullets from $pr_list (Step 2), and close the final subsection with the Full changelog: line using the blob link from references/release-notes-format.md. Use Edit.

   If a generated bookkeeping subsection is already present under this heading, replace it in place rather than appending a second copy. A re-run reaching this branch is ordinary: the Step 9 tag gate sits after the release commit, so declining it and running again later brings the run back to its own changelog section.

3. It does not exist — generate, then append. Dispatch one general-purpose agent to write the narrative ## v{parent target} section. The skill already holds this context — paste the actual values into the prompt, do not tell the agent to go read them:

   • The version plan from Step 3: parent baseline → target, and for each changed/new child its name, baseline → target, level, and new/changed status.
   • The commit log git log {range} --oneline and git diff {range} --stat, plus, per changed plugin, git diff {range} --stat -- {source}/ so the agent can attribute each change to its plugin, plus a suite-level stat git diff {range} --stat -- docs/ README.md CONTRIBUTING.md CHANGELOG.md .claude-plugin/ labeled as the evidence for the ### han parent section (repo-root changes outside any plugin directory).
   • $pr_list (PR numbers, titles, authors).
   • $issue_list (Step 2): each closed issue's number, title, opener, contributors, closing PR(s), and the relevant fix, so the narrative can credit the issue opener where it describes that fix.
   • $release_context from Step 1 (may be empty).
   • The two newest existing ## v{X.Y.Z} sections from CHANGELOG.md verbatim, as the register model.
   • The "Register and voice" and "Per-plugin structure" constraints from references/changelog-rules.md, pasted in full.

   Prompt the agent to: produce only the markdown for the ## v{parent target} section — a one-paragraph summary that names the parent's new version and lists each changed/new child with its version, then one ### {plugin} v{version} sub-heading per changed or new plugin (parent first), each describing only that plugin's changes (using #### for topic subsections when needed), then a release-level ### Deferred (YAGNI) subsection only when work was deliberately cut; when a change closes a tracked issue, name the fix and credit the issue opener inline as a profile link [@{login}](https://github.com/{login}); render every @mention as a [@{login}](https://github.com/{login}) profile link, never flat text; match the register of the two pasted sections; obey every hard voice constraint; attribute every change to the plugin whose directory it touched; never invent changes not present in the commits, diff, PR list, or issue list; return only the section markdown with no preamble. If the agent returns anything else, discard it and re-issue with an explicit "return only the section markdown" reminder.

   Insert the returned section directly under the # Han Release Notes title, above the previous newest entry (Edit/Write). Then append the generated bookkeeping subsections to it exactly as in the augment case.

Step 6: Assemble the release notes body

Build the GitHub release body per references/release-notes-format.md: the release's summary paragraph first (the one-paragraph overview from the ## v{parent target} narrative, with no heading above it), then ## What's Changed and the PR lines (* {title} by @{login} in {url}, newest merge last), then an ## Issues closed section built from $issue_list (omitted when empty), then every ### {plugin} v{version} sub-heading of the narrative excluding the ## v{parent target} heading itself and the generated PR/commits/issues bookkeeping subsections (the summary paragraph already leads the body, so it is not repeated here), then the **Full changelog:** blob link and the **Full Changelog:** compare link (compare line omitted on a first release). Compute the blob anchor by lowercasing v{parent target} and deleting every character that is not a-z, 0-9, or - (v3.0.0 → v300). Write the assembled body to /tmp/han-release-notes-v{parent target}.md with the Write tool. Do not assemble it with shell echo/printf.

Step 7: Show the prepared release

Print to the operator, regardless of mode:

• The full version plan: the release tag {parent plugin name}--v{parent target}, the parent's baseline → target and how it was decided (ahead-of-tag → used as-is, or computed-and-confirmed at Step 3), and one line per child (bumped baseline → target, unchanged at version, or new at version).
• The branch the tags will point at, plus the non-default-branch note from Step 1.3 if it applies.
• Any non-blocking advisory from Step 3 (under-bump warning on any plugin) and the post-release advisory check: if CLAUDE.md states a "Current version:" that does not equal the parent target, note it as a follow-up the operator may want to make (do not edit CLAUDE.md; it is out of scope).
• The exact CHANGELOG diff for the ## v{parent target} section.
• The full assembled release notes body from Step 6.
• The publish mode: published --latest, or draft (only if draft_release).

If pause_before_publish is true: use AskUserQuestion (header: "Publish release") — options: publish now (proceed to Step 8), abort (stop, having changed only local files). Do not push or publish until approved.

If pause_before_publish is false (default): continue to Step 8 without pausing.

This gate is opt-in and covers the whole release. It is separate from the Step 9 tag gate, which always fires and covers the tags only. When both fire they arrive close together; that is intended, and the distinct header values are what tell them apart.

Step 8: Commit and push the release commit

The operator's request to tag and publish authorizes the commit and push required to do it.

1. Commit the release prep. Stage CHANGELOG.md, .claude-plugin/marketplace.json, and every {source}/.claude-plugin/plugin.json that Step 4 changed. Commit with chore(release): v{parent target}. The commit subject keeps the plain version; it is a message, not a ref. If nothing is staged (augment produced no diff and no version changed — unlikely), skip the commit and note it.

2. Record {release commit}. git rev-parse HEAD. Every tag must point here, and Step 9 re-checks it.

3. Push the commit, before any tag. git push origin HEAD. If this fails, stop: no tag has been created, so nothing irreversible has happened. Pushing a tag first would transfer the commit to GitHub without putting it on any branch, and a tag can never be moved afterwards.

Step 9: Approve the tags (mandatory gate)

The commit is on GitHub and no tag exists yet. Everything past this point is irreversible.

1. Compute each plugin's tag name as {name}--v{target}, for every plugin in plugins.

2. Classify them against the remote in one call:

   ${CLAUDE_SKILL_DIR}/scripts/remote-tag-state.sh {release commit} {tag name}...
   

   It prints {tag}\t{state}\t{sha} per tag. Exit 2 means the remote could not be read: stop, because a run that cannot see the remote must not guess. The four states and what each one means for the walk:

   State	What Step 10 does with it
   absent	Create and push it.
   remote-at-commit	Skip it. Already published at this commit.
   remote-at-other-commit	Stop the run now (see 3 below).
   local-only	Push it. This is never a skip.

3. Any remote-at-other-commit stops the run here, before a single tag is created. Name the tag, the commit it points at, and {release commit}. Say plainly that pushing is not the recovery: the push is rejected identically every time, GitHub does not allow a published release's tag to be moved or deleted, and this skill never forces one. The way out is a different version number, which is free right now and costs a burned version once the walk starts.

4. Ask for approval with AskUserQuestion (header: "Tag plugins"). Show:

   • The tags to create, and separately the tags already on GitHub.
   • {release commit} and the branch it is on, with the Step 1.3 note if it is not the default branch.
   • When draft_release is true, that the draft flag holds back the release page only. Every tag is still created and pushed, and none of them can be moved afterwards.

   Options: create and push the tags (recommended, first); abort.

5. This gate always fires. Unlike Step 3d it has no skip condition, because the tags are permanent whether or not a version needed computing.

6. Treat an empty or unreadable answer as an abort. Do not read it as approval. AskUserQuestion has a known failure mode that returns empty answers silently, which is why it is absent from allowed-tools (see the note under the frontmatter). This is the one gate in this skill where that failure would be unrecoverable.

7. On abort: stop. Nothing is tagged and nothing is published. The release commit is already made and pushed, so say so, name it, and say that re-running later picks up from there: the changelog section already exists, so the run will augment it rather than regenerate it.

8. Re-read git rev-parse HEAD after approval. If it no longer equals {release commit}, stop. The tags are created against the current checkout, so a commit made while the gate was waiting would tag something that was never pushed.

Step 10: Tag every plugin

Walk plugins, taking the entry whose name equals parent plugin name first, then the rest in listed order. The parent goes first so the tag the GitHub release attaches to is the one most likely to exist if the walk stops partway. Order by name, not by position in the file.

For each plugin, act on its Step 9 state:

• remote-at-commit — skip. Record it as already on GitHub.

• local-only — git push origin refs/tags/{tag name}. Record it as newly pushed.

• absent — create and push it in one call:

  command claude plugin tag {source} --push
  

  The command builtin is required, not stylistic: a bare claude runs whatever function or alias the operator's shell defines, which commonly blocks waiting for terminal input. The command derives the tag name from the plugin's manifest and marketplace entry and re-checks that the two agree, so a disagreement that reached this point surfaces here.

Never pass --force. It skips both the dirty-tree and tag-exists checks, and this skill does not move tags.

Reading a failure. The command exits 1 for both a benign refusal and a real failure, so the exit status alone does not tell them apart. Match the message:

• already exists locally — the tag is on this machine but was not on the remote at Step 9, so a push is what is needed. Push it as in the local-only case above. Do not re-run the tagging command; it will refuse identically.
• Any other non-zero exit stops the run. That includes a push rejection, an uncommitted change under a plugin folder, and a version disagreement that Step 3.5 should have caught. A stop here is unexpected rather than routine.

On a stop mid-walk, do not publish. Report the full tag state: which tags are on GitHub, which exist only on this machine, and for each of the latter the literal recovery command git push origin refs/tags/{tag name}. Re-running /han-release does not retry a failed push, so never offer that as the recovery.

Step 11: Publish the GitHub release

1. Re-run the classification from Step 9 over every plugin's tag name. Every one must read remote-at-commit. If any does not, stop and report as in Step 10. Publishing with a tag missing from GitHub is the failure this gate exists to prevent.

2. Publish, per references/release-notes-format.md, using /tmp/han-release-notes-v{parent target}.md (the filename keeps the plain version; it is a local path, not a ref):

   • No release exists for the tag (gh release view {parent plugin name}--v{parent target} fails): gh release create {parent plugin name}--v{parent target} --verify-tag --title "v{parent target}" --notes-file /tmp/han-release-notes-v{parent target}.md plus --latest, plus --draft only when draft_release is true (never --latest together with --draft).

     --verify-tag is required. Without it gh silently creates a missing tag at the default branch's head, which would mint a permanent tag at a commit nobody released. The --title keeps the plain v{parent target} so the releases page reads continuously across the naming change.

   • A release already exists: do not create a second one. gh release edit {parent plugin name}--v{parent target} --notes-file /tmp/han-release-notes-v{parent target}.md. Add --draft=false only when the operator asked to publish an existing draft (and draft_release is not set). Report it as updated, not created.

Step 12: Report

Report concisely:

• The full version plan (parent and each child, and how the parent version was decided).
• {release commit}, and whether it is on the default branch.
• One line per plugin naming its final tag state: newly pushed, already on GitHub, or present only on this machine. Those three are different, and "created" plus "skipped" cannot express the third, which is the state a partly-failed release leaves behind.
• Any tag still local-only, with its literal git push origin refs/tags/{tag name} recovery line.
• The files committed and the commit; the release URL (or draft URL); whether the CHANGELOG section was augmented or generated.
• Any advisories from Step 7 (under-bump, CLAUDE.md version drift, non-default release branch).

If the operator aborted at Step 7 or Step 9, report exactly what was changed locally, what was pushed, and what was not.