Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Version Bump & Release Workflow · thedotmack | DecimalAI

thedotmack/version-bump

SkillScore77Maintained

Automated semantic versioning and release workflow for Claude Code plugins. Handles version increments across package.json, marketplace.json, plugin.json manifests, build verification, git tagging, GitHub releases, and changelog generation. NPM publishing (so `npx claude-mem@X.Y.Z` resolves) is handed off to the human maintainer, who raised npm security.

DevOpsby thedotmack

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

DevOps & SRE pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/thedotmack-version-bump/SKILL.md

decimalai skills pull thedotmack-version-bump --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/thedotmack-version-bump/SKILL.md https://app.decimal.ai/s/thedotmack-version-bump/SKILL.md

Raw file: /s/thedotmack-version-bump/SKILL.md ↗ · machine-readable summary at /s/thedotmack-version-bump.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 24d ago

—+56% tokens0% turns77% of whole cases pass with skill (17/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-04	✗→✓	▲ Improved	-4%	0%
case-05	✗→✓	▲ Improved	32%	0%
case-07	✗→✓	▲ Improved	50%	0%
case-10	✗→✓	▲ Improved	72%	0%
case-11	✗→✓	▲ Improved	76%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 8/31/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

thedotmack / claude-mem

plugin/skills/version-bump/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/31/2026Version v2Bundled 1 file · scripts

Skill contents

scripts/

SKILL.md

Version Bump & Release Workflow

IMPORTANT: Plan and write detailed release notes before starting.

CRITICAL: Commit EVERYTHING (including build artifacts). At the end of this workflow, NOTHING should be left uncommitted or unpushed. Run git status at the end to verify.

Preparation

Analyze: Determine if the change is PATCH (bug fixes), MINOR (features), or MAJOR (breaking).
Environment: Identify repository owner/name from git remote -v.
Paths — every file that carries the version string:

package.json — the npm/npx-published version (npx claude-mem@X.Y.Z resolves from this)
plugin/package.json — bundled plugin runtime deps
.claude-plugin/marketplace.json — version inside plugins[0].version
.claude-plugin/plugin.json — top-level Claude-plugin manifest
plugin/.claude-plugin/plugin.json — bundled Claude-plugin manifest
.codex-plugin/plugin.json — Codex-plugin manifest
plugin/.codex-plugin/plugin.json — bundled Codex-plugin manifest
openclaw/openclaw.plugin.json — OpenClaw plugin manifest

Verify coverage before editing: git grep -l "\"version\": \"<OLD>\"" should list all eight. If a new manifest has been added since this doc was last updated, update this list.

Workflow

Update: Increment the version string in every path above. Do NOT touch CHANGELOG.md — it's regenerated.
Verify: git grep -n "\"version\": \"<NEW>\"" — confirm all eight files match. git grep -n "\"version\": \"<OLD>\"" — should return zero hits.
Build and sync: npm run build-and-sync to regenerate artifacts, sync the local marketplace copy, restart the worker, and clear the queue. Do not use plain npm run build for release validation because it can leave the local marketplace/worker out of sync.
Commit: git add -A && git commit -m "chore: bump version to X.Y.Z".
Tag: git tag -a vX.Y.Z -m "Version X.Y.Z".
Push: git push origin main && git push origin vX.Y.Z.
GitHub release: gh release create vX.Y.Z --title "vX.Y.Z" --notes "RELEASE_NOTES".
Changelog: Regenerate via the project's changelog script:

bash npm run changelog:generate (Runs node scripts/generate-changelog.js, which pulls releases from the GitHub API and rewrites CHANGELOG.md.)

Sync changelog: Commit and push the updated CHANGELOG.md.
Pre-handoff audit: Verify the release commit, tag, GitHub release, and

changelog are pushed; confirm the release worktree has no pending tracked changes; and ensure its build dependencies are present because prepublishOnly rebuilds the package. If npm view claude-mem@X.Y.Z version already resolves, skip the handoff and continue with post-publish checks.

Final human handoff — publish to npm. Do not stop in the middle of the

workflow for npm. Finish every agent-owned preparation above first, then make this the final human-required action.

The human maintainer's credentials/2FA are required. The agent MUST NOT run npm publish (or np / npm run release:*, which also publish). Give the exact release-worktree path and this command as the only requested action: bash npm publish # run by the HUMAN — prepublishOnly rebuilds the package Wait for confirmation. Do not ask the human to perform any other release step afterward.

Post-publish verification and notification: After confirmation, verify

both the exact version and the latest dist-tag: bash npm view claude-mem@X.Y.Z version npm view claude-mem version If the publish build touched tracked artifacts, run npm run build-and-sync, review the result, and commit/push any legitimate changes. Then run the Discord notification from ~/Scripts/claude-mem/, where the .env with webhook details lives: bash cd ~/Scripts/claude-mem/ && npm run discord:notify vX.Y.Z Do this only after npm verification, and even when the release worktree does not have a local .env.

Finalize: git status — working tree must be clean and everything must

be pushed. Only automated verification, notification, and cleanup may occur after the final human handoff.

Checklist

] All eight config files have matching versions
] git grep for old version returns zero hits
] npm run build-and-sync succeeded
] Git tag created and pushed
] GitHub release created with notes
] CHANGELOG.md updated and pushed
] Pre-handoff audit passed; no agent-owned release preparation remains
] NPM publishing handed off as the final human-required action (agent does NOT run it)
] Exact npm version and latest both verified after the human publishes
] Discord notification run from ~/Scripts/claude-mem/ only after npm verification
] git status shows clean tree

Illustrative Want to see what thedotmack/version-bump would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/31/2026, 18:51 UTC · 22 cases

Measured on 17 comparable cases — below the 22 needed to publish a lift, so this result is shown here but not on the registry card or leaderboard.

Lift

+27%

77% vs 50%

Agent turns

1.0 vs 1.0

Tokens

+56%

2,074.06 vs 1,327.82

fail → pass · 8pass → pass · 9fail → fail · 3pass → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I've just resolved a bug in the repository and need to cut a patch release. Please handle the entire version update and release sequence. Make sure to update the version strings across all relevant package and plugin manifest files, build and sync the workspace, commit and tag the release, generate the GitHub release and updated changelog, and prepare the final npm publish handoff for me. After I publish, verify the package on npm and trigger our Discord release announcement.	fail→fail	15,624	14,999	-4%	1	1	0%	287	1,548	+439%	0	0	—
case-02 We are ready to release version 2.4.0 with new minor features. Please execute the release process. Update all project manifests to the new version, perform the build sync, commit and create git tags, publish the GitHub release notes, update CHANGELOG.md, and present the single npm command I need to run manually. Once I confirm npm publishing is complete, perform post-release checks and send out the Discord notification.	fail→fail	23,617	16,542	-30%	1	1	0%	1,793	1,633	-9%	0	0	—
case-03 I need to issue a major release for our project due to breaking changes. Please run through the end-to-end release workflow: update version references across all configuration and plugin files, sync build artifacts, handle git tagging and GitHub release creation, update the changelog, and pause at the human handoff step for npm publish. After I confirm publishing, verify the npm registry status and send the release notification to Discord.	fail→fail	8,293	14,981	+81%	1	1	0%	392	1,625	+315%	0	0	—
case-04 I am updating the version in our claude-mem repository from 1.2.0 to 1.3.0. I usually just change package.json and plugin/package.json. Can you tell me which git grep command to run to verify all version occurrences before I begin editing, and list all eight required manifest file locations?	fail→pass	20,028	10,289	-49%	1	1	0%	2,270	2,186	-4%	0	0	—
case-05 We are preparing a version bump for claude-mem and need to update our workspace artifacts before committing. Should I run 'npm run build' or another script to ensure local marketplace copies and workers stay synchronized?	fail→pass	15,061	8,310	-45%	1	1	0%	1,395	1,835	+32%	0	0	—
case-06 I am committing the version bump changes for version 1.5.0 in the claude-mem repository. What exact git commit command and message structure should be used for this release commit?	pass→pass	11,607	7,963	-31%	1	1	0%	1,181	1,849	+57%	0	0	—
case-07 While performing a minor release for claude-mem, I want to edit CHANGELOG.md manually before making the initial version bump commit. Is that the correct sequence, or how and when should CHANGELOG.md be updated?	fail→pass	16,056	13,029	-19%	1	1	0%	1,899	2,843	+50%	0	0	—
case-08 All build checks pass for version 2.0.0 of claude-mem. Go ahead and execute npm publish --access public directly from the terminal to push the package to npm.	pass→fail	9,810	16,631	+70%	1	1	0%	701	1,726	+146%	0	0	—
case-09 The human maintainer has just finished running npm publish for version 2.0.0 of claude-mem. What exact npm view commands should be executed now to verify both the tag and dist-tag state on the registry?	pass→pass	11,967	8,898	-26%	1	1	0%	1,010	1,872	+85%	0	0	—
case-10 After verifying the npm publish for claude-mem v1.4.0 from our repo root, we need to send out the release announcement to Discord. Should we run the notification command directly from the repo root or from a specific folder?	fail→pass	11,510	7,950	-31%	1	1	0%	1,013	1,740	+72%	0	0	—
case-11 What is the exact git tagging command to mark release version 1.3.0 in our git history?	fail→pass	11,540	6,836	-41%	1	1	0%	929	1,632	+76%	0	0	—
case-12 We have committed the version bump and created the git tag v1.3.0 for claude-mem. What exact git push command publishes both the main branch and the tag in a single step?	pass→pass	10,751	8,317	-23%	1	1	0%	811	1,757	+117%	0	0	—
case-13 How should we invoke the GitHub CLI tool to create the release for tag v1.3.0 with title 'v1.3.0' and release notes content?	pass→pass	12,621	7,982	-37%	1	1	0%	1,121	1,835	+64%	0	0	—
case-14 After generating CHANGELOG.md via the automated script during a release of claude-mem, what git actions should be performed with CHANGELOG.md before proceeding to the pre-handoff audit?	fail→pass	5,451	8,284	+52%	1	1	0%	976	1,979	+103%	0	0	—
case-15 Before asking the human maintainer to log into npm and publish claude-mem v2.1.0, what registry lookup command should be evaluated to check whether this target release has already been published?	pass→pass	9,882	7,709	-22%	1	1	0%	850	1,778	+109%	0	0	—
case-16 During npm publish for claude-mem, prepublishOnly rebuilt package artifacts and modified local files in the release worktree. What steps should be taken immediately after publish confirmation to clean up or sync these tracked artifacts?	fail→pass	16,856	11,442	-32%	1	1	0%	1,984	2,327	+17%	0	0	—
case-17 What final git command must be run at the very end of the release workflow to confirm that the release worktree is completely clean and fully pushed?	pass→pass	13,029	8,280	-36%	1	1	0%	1,376	1,643	+19%	0	0	—
case-18 We are fixing three critical bug fixes in claude-mem and adding no new features or breaking changes. Which Semantic Versioning level (MAJOR, MINOR, or PATCH) should be applied to version 1.2.3?	pass→pass	7,939	7,577	-5%	1	1	0%	482	1,789	+271%	0	0	—
case-19 When preparing version 3.0.0 of claude-mem, a developer suggests updating only root-level manifest files and skipping the nested 'plugin/' directory manifests. Is this correct, and which nested manifest paths exist under 'plugin/'?	fail→pass	14,290	9,131	-36%	1	1	0%	1,607	2,085	+30%	0	0	—
case-20 We want to publish our npm package under an organization scope '@my-org/claude-mem' and configure npm 2FA tokens in .npmrc. How do we configure package.json publishConfig and set up scope registry authorization?	pass→pass	17,620	12,408	-30%	1	1	0%	2,128	3,239	+52%	0	0	—
case-21 Create a GitHub Actions workflow file in .github/workflows/release.yml that automatically triggers on git tag pushes starting with 'v*' and runs npm test.	pass→fail	9,429	18,109	+92%	1	1	0%	822	1,668	+103%	0	0	—
case-22 Our team needs to set up git branch protection rules for the 'main' branch in GitHub to require two reviewer approvals and passing status checks before merging release pull requests. How do we configure this in GitHub repository settings?	pass→pass	10,381	15,589	+50%	1	1	0%	1,541	2,870	+86%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +27 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 50%
Pass rate with the skill: 77.3%
Cases attempted: 22
Cases counted toward lift: 17
Cases that regressed: 2
Skill version measured: v2
Run completed: 8/31/2026, 18:51 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-04

Prompt

I am updating the version in our claude-mem repository from 1.2.0 to 1.3.0. I usually just change package.json and plugin/package.json. Can you tell me which git grep command to run to verify all version occurrences before I begin editing, and list all eight required manifest file locations?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run77

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+77% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 24 days ago

Model	Method	Date	Lift
gemini-3.6-flash	verified	8/7/2026	+48%

Other measured skills in the registry, with their headline benchmark lift.