Scaffolding Maintainer logo

Scaffolding Maintainer

Community
hmohamed01
scaffolding-maintainer

Maintains and improves the project-scaffolding skill. This skill should be used when the user asks to "update the scaffolding skill", "refresh framework versions", "are the versions stale", "check for outdated packages", "improve project-scaffolding", "act on the scaffolding observations", "add a project type to the scaffolder", "fix a scaffolding template", or after a scaffolding session logged an entry in assets/observations.md. Handles two distinct jobs: refreshing dependency facts from the npm and PyPI registries, and patching the skill's procedures, templates, and wizard options based on recorded real-world usage.

Overview

Publisherhmohamed01
RepositoryClaude-Code-Scaffolding-Skill
Skill namescaffolding-maintainer
Stars
56
Forks
4
Bundled files
3
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by hmohamed01 on GitHub. Read the source before you install it.

Installation

Install the Scaffolding Maintainer AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/hmohamed01/Claude-Code-Scaffolding-Skill/tree/main/scaffolding-maintainer
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/hmohamed01/Claude-Code-Scaffolding-Skill.git /tmp/Claude-Code-Scaffolding-Skill
mkdir -p .claude/skills
cp -r /tmp/Claude-Code-Scaffolding-Skill/scaffolding-maintainer .claude/skills/scaffolding-maintainer
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Scaffolding Maintainer in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Scaffolding Maintainer on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Scaffolding Maintainer is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Scaffolding Maintainer

Improves the project-scaffolding skill along two separate paths. Keep them separate — they have different sources of truth and different failure modes.

PathWhat changesSource of truthRisk if wrong
A. Fact refreshVersion numbers in assets/versions.jsonnpm / PyPI registriesGenerated projects pin versions that do not exist or do not install
B. Procedure improvementSKILL.md, references, wizard options, templates in scaffold.pyassets/observations.mdThe wizard recommends something wrong or generates broken code

Absolute Rules

  1. Never write a version number from memory. Model recall of package versions is unreliable and silently wrong. Every version in versions.json must come from refresh_versions.py, which queries the registries directly. If the network is unavailable, stop and say so.
  2. Never auto-apply a major version bump. A major bump can invalidate the template source in scaffold.py, not just the number. Follow references/template-drift.md first.
  3. Never finish without running verification. scripts/verify_scaffold.py must pass before repackaging, and --install must pass before adopting any version change. A version that resolves is not a version that builds.
  4. Never delete observation entries. Mark them resolved in place so the log stays auditable.

Path A: Refresh Versions

Run from the project-scaffolding/ directory.

bash
python3 scripts/scaffold.py --version-status      # how stale is it?
python3 scripts/refresh_versions.py               # dry run; reports -> bump  !! review

Read the report:

  • -> bump lines are minor/patch moves within the existing major. Because the manifest uses caret (^) and >= ranges, users already resolve to these versions at install time. Applying them is safe and mostly cosmetic.
  • !! review lines are major bumps, pre-1.0 minors, or 0.0.x patches. These need judgment. Do not apply them with --include-review until the template check in references/template-drift.md is done for each affected package.
  • A !! line ending in [safe now: X] has an in-series update available. --apply takes those automatically: the major stays held, but the package still gets its fixes. An in-series bump is safe for the package itself, not necessarily for its peers — bumping TypeScript 5.8 to 5.9 once broke Angular 19, whose peer range capped at <5.9. Always run --install afterwards.

Apply the safe set, then handle reviews individually:

bash
python3 scripts/refresh_versions.py --apply       # applies bump lines only
python3 scripts/verify_scaffold.py                # regenerate every project type

When a major bump is genuinely not wanted — a package pinned for compatibility with something else, such as React Native tracking an Expo SDK — record it in the manifest's hold map with a reason rather than repeatedly declining it:

json
"hold": { "react-native": "must match the Expo SDK version the wizard offers" }

Path B: Act on Observations

Read assets/observations.md. Group entries by tag and decide what each class warrants:

TagTypical fix
default-rejectedIf one default is rejected repeatedly, change the recommendation in SKILL.md and references/wizard-options.md — or present it as an explicit choice instead of a default
missing-optionAdd the framework or option: update references/wizard-options.md, SKILL.md's type list, and implement a creator method in scaffold.py
broken-outputFix the generator, then add the case to verify_scaffold.py so it cannot regress
template-driftUpdate the template method in scaffold.py to match the pinned major

Require a pattern before changing a default. A single rejection is one user's preference, not evidence. Two or three independent entries with the same shape justify a change. State the evidence count when proposing one.

When an entry is acted on, append -> RESOLVED <date>: <what changed> to that line. When an entry is judged not actionable, append -> WONTFIX <date>: <why>.

See references/review-workflow.md for the full patching procedure, including how to add a new project type end to end.

Verification (mandatory)

bash
python3 scripts/verify_scaffold.py                # all types generate, configs parse
python3 scripts/verify_scaffold.py --snapshot     # record current output as the golden baseline
python3 scripts/verify_scaffold.py --compare      # diff against the golden baseline

Use --compare after any change that should not alter output (a refactor). Use --snapshot to re-baseline deliberately after a change that should. An unexplained diff means stop and investigate, not re-snapshot.

--install is the mode that finds real defects. Generation checks confirm files are well-formed; they cannot tell that a template imports a package nobody declared, that a lint script points at a removed command, or that a tsconfig option was deprecated. Only installing the project and running its scripts finds those. Run it after any version adoption, and after any template change:

bash
python3 scaffolding-maintainer/scripts/verify_scaffold.py --install --only react nextjs

Node version matters: run it on the Node major the manifest targets, since several toolchains refuse to start on older releases.

Stage the baseline with -f. Generated projects contain their own .gitignore files, and a nested .gitignore overrides the repository's rules for its subtree — so a plain git add drops files (.vscode/settings.json, src/lib/*) from the baseline without reporting anything:

bash
python3 scaffolding-maintainer/scripts/verify_scaffold.py --snapshot
git add -f tests/golden
git diff --cached --name-only -- tests/golden | wc -l   # must equal the file count on disk

Already-tracked files stay tracked regardless of ignore rules; -f matters only for files a snapshot newly introduces.

Repackaging

After any change under project-scaffolding/, repackage from the repository root:

bash
cd project-scaffolding && zip -r ../project-scaffolding.skill SKILL.md scripts/ references/ assets/
cd ../scaffolding-maintainer && zip -r ../scaffolding-maintainer.skill SKILL.md scripts/ references/

Note assets/ is included — the version manifest lives there, and a package without it will fail at import with a missing-manifest error.

Then record the change in CHANGELOG.md at the repository root.

Additional Resources

  • references/template-drift.md — how to check whether a major bump invalidates template source, package by package
  • references/review-workflow.md — full procedure for patching the skill, including adding a new project type
  • scripts/verify_scaffold.py — generates every project type and validates the output

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Scaffolding Maintainer AI skill do?

Maintains and improves the project-scaffolding skill. This skill should be used when the user asks to "update the scaffolding skill", "refresh framework versions", "are the versions stale", "check for outdated packages", "improve project-scaffolding", "act on the scaffolding observations", "add a project type to the scaffolder", "fix a scaffolding template", or after a scaffolding session logged an entry in assets/observations.md. Handles two distinct jobs: refreshing dependency facts from the npm and PyPI registries, and patching the skill's procedures, templates, and wizard options based...

Why use Scaffolding Maintainer on TypingMind?

Because you install it once and use it with any model. Scaffolding Maintainer is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Scaffolding Maintainer in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/hmohamed01/Claude-Code-Scaffolding-Skill/tree/main/scaffolding-maintainer. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Scaffolding Maintainer?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Scaffolding Maintainer?

As many as you like. As long as a model supports skills, you can use Scaffolding Maintainer with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Scaffolding Maintainer AI skill free?

It is published on GitHub by hmohamed01. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇