HyperFrames by HeyGen
Official HyperFrames by HeyGen: the best AI video editor in chat. Captions, motion graphics, MP4.
Interactive UI for local node modules inspection
If you made this, paste this code into your README. The badge links back to this page.
[](https://mcp.popnx.com/en/skills/antfu-node-modules-inspector)MCP servers that connect the AI to services for the same work
Official HyperFrames by HeyGen: the best AI video editor in chat. Captions, motion graphics, MP4.
draw.io diagrams your AI agent draws on a live browser canvas that you can also edit by hand
Create, inspect, validate, and save editable 3D building scenes with Pascal's MCP server.
Same kind of work, ordered like the Skills list
| This one node-modules-inspector | impeccable | diagram-design | frontend-slides | guizang-ppt-skill | |
|---|---|---|---|---|---|
| Maker | antfu | pbakaus | cathrynlavery | zarazhangrui | op7418 |
| Skills | 1 | 1 | 1 | 1 | 1 |
| GitHub stars | 2,962 | 111,265 | 67,613 | 32,407 | 29,891 |
| Last code change | Sep 17, 2026 | Nov 17, 2026 | Nov 13, 2026 | Jun 24, 2026 | Oct 30, 2026 |
| License | MIT | Apache-2.0 | MIT | MIT | AGPL-3.0 |
1 skills (instruction sets) by antfu that an AI reads and follows · not an MCP server
MCP = the hands
Connects the AI to an outside service or data, such as reading Notion or creating an invoice. A program or server has to be running.
Skill = the manual
A text file that teaches the AI how to do a task. It connects to nothing.
Some overlap: a few Skills run scripts that call tools, and a few MCP servers only send knowledge. Start from the task you want done.
Claude Code
From the maker’s README and the folders in the repo
GitHub account antfu (person), 11 years old
★ 2,962 stars · code last changed Sep 17, 2026
No, plain text files the AI reads and follows
No scripts, so it does not go online by itself
No hooks or writes to machine settings found
Checked automatically from the repo files at commit 602140f (read only, nothing was run). Not a safety guarantee: read it and decide for yourself
First: which AI do you use?
Pick your app and this page shows only how to install it there
Uploading to claude.ai would fail: the description is over 200 characters (724). Use "Let the AI install it" instead
Copy this sentence and paste it into a Claude Code chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Restart Claude Code, then type /node-modules-inspector or ask for a matching task and Claude picks it up
Checked against official docs Nov 12, 2026 · Source: code.claude.com
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
git clone --depth 1 https://github.com/antfu/node-modules-inspector.git /tmp/popnx-skill-antfu-node-modules-inspector-602140f && mkdir -p ~/.claude/skills && cp -R '/tmp/popnx-skill-antfu-node-modules-inspector-602140f/skills/node-modules-inspector' ~/.claude/skills/node-modules-inspectorThe ChatGPT desktop app has Codex built in, and Codex reads skills from ~/.agents/skills
Open Codex in the ChatGPT desktop app and paste this sentence into the chat
The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Put the node-modules-inspector folder at ~/.agents/skills/node-modules-inspector
~ is your home folder ($HOME)
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
git clone --depth 1 https://github.com/antfu/node-modules-inspector.git /tmp/popnx-skill-antfu-node-modules-inspector-602140f && mkdir -p ~/.agents/skills && cp -R '/tmp/popnx-skill-antfu-node-modules-inspector-602140f/skills/node-modules-inspector' ~/.agents/skills/node-modules-inspectorChecked against official docs Nov 15, 2026 · Source: learn.chatgpt.com
Skills you install yourself do not work here; only skills that come with a plugin from the directory
To use this skill with ChatGPT, use the ChatGPT desktop app as above
Checked against official docs Nov 15, 2026 · Source: learn.chatgpt.com
Copy this sentence and paste it into a Codex chat · Works in Codex inside the ChatGPT desktop app, Codex CLI and the IDE extension
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Codex picks up newly installed skills by itself; if one does not show up, restart Codex · To use it: in ChatGPT type @ and pick the skill · in Codex CLI or the IDE extension run /skills or type $ and the skill name
Checked against official docs Nov 15, 2026 · Source: learn.chatgpt.com
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
git clone --depth 1 https://github.com/antfu/node-modules-inspector.git /tmp/popnx-skill-antfu-node-modules-inspector-602140f && mkdir -p ~/.agents/skills && cp -R '/tmp/popnx-skill-antfu-node-modules-inspector-602140f/skills/node-modules-inspector' ~/.agents/skills/node-modules-inspectorCopy this sentence and paste it into a Cursor chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
In Cursor’s Agent chat, type / and pick node-modules-inspector
Checked against official docs Nov 12, 2026 · Source: cursor.com
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
git clone --depth 1 https://github.com/antfu/node-modules-inspector.git /tmp/popnx-skill-antfu-node-modules-inspector-602140f && mkdir -p ~/.cursor/skills && cp -R '/tmp/popnx-skill-antfu-node-modules-inspector-602140f/skills/node-modules-inspector' ~/.cursor/skills/node-modules-inspectorFor other apps that read skills. The sentence gives the paths for Claude Code, Codex and Cursor so the AI picks the one for your app · Copy this sentence into the app’s chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Checked against official docs Nov 12, 2026 · Source: code.claude.com, learn.chatgpt.com, cursor.com
Other apps that support Agent Skills (such as Gemini CLI or GitHub Copilot) use different folders. Check that app’s guide, then copy the folder with the SKILL.md there.
Skill folders checked against the official Claude Code, Codex and Cursor guides on Nov 11, 2026
✓ Commands checked Nov 17, 2026
The text below comes from the maker’s repo, not from PopNx MCP
Showing the first 8,000 characters · Open the full file on GitHub
---
name: node-modules-inspector
description: >
Inspects a project's installed node_modules and produces three reports:
duplicated packages (installed in multiple versions), packages sorted by
install size, and maintenance actions (dep-upgrade opportunities + publint
findings, grouped by consumer/author). Use when the user wants to audit
dependencies, find duplicate packages, check what's taking up disk space in
node_modules, identify outdated peer/prod dependencies that newer dependents
could upgrade past, or list publint problems. Available as a CLI
(`npx node-modules-inspector report <duplicates|sizes|maintainers> [--json]`)
or an MCP stdio server (`npx node-modules-inspector mcp`) exposing the same
three reports as agent tools. Works with pnpm, npm, and bun.
---
# node-modules-inspector
`node-modules-inspector` is a CLI + MCP server that inspects the installed `node_modules` of the current project and produces structured reports. Three reports, same underlying analysis pipeline:
| Report | Answers |
|---|---|
| `duplicates` | Which packages are installed in multiple versions? |
| `sizes` | Which packages take up the most disk space? |
| `maintainers` | Which consumers have dep-upgrade opportunities or publint issues, grouped by package and author? |
Reports run against the real on-disk `node_modules` — no registry calls are required for the basic shape; npm metadata is fetched only to enrich the maintainers report (gated by config).
Works with pnpm, npm, and bun. The default `npx node-modules-inspector` (with no subcommand) opens a Vue web UI for humans; agents should use the `report` and `mcp` subcommands below.
## When to reach for this
Trigger on any of:
- "audit my dependencies", "find duplicate packages", "node_modules cleanup"
- "what's taking up disk space in node_modules"
- "which deps are outdated" / "what dep-upgrade opportunities are there"
- "show me publint issues across my deps"
- "who maintains my dependencies"
Don't reach for it for: registry-only questions (use `fast-npm-meta`), bundle-size analysis of a single package (use a bundler-specific tool), security audits (use `npm audit` / `osv-scanner`).
## CLI mode
All subcommands share these options:
- `--root <dir>` — project root (default: cwd)
- `--config <file>` — config file (default: `node-modules-inspector.config.{ts,js,json}`)
- `--depth <n>` — max dependency depth to traverse (default: `8`)
- `--json` — emit JSON to stdout; pretty ANSI table otherwise
Progress logs always go to stderr, so `... --json` is pipe-safe.
### duplicates
```sh
npx node-modules-inspector report duplicates --json
```
Options:
- `--min-versions <n>` — only include packages installed at this many versions or more (default: `2`)
- `--limit <n>` — cap result count
Output shape:
```json
[
{
"name": "@typescript-eslint/scope-manager",
"versions": ["8.56.1", "8.59.1", "8.59.2", "8.59.4"],
"specs": ["@typescript-eslint/scope-manager@8.56.1", "..."]
}
]
```
Versions are sorted ascending by semver. Entries are sorted by version-count descending. Use this to find dedupe targets — `pnpm dedupe` / `npm dedupe` resolves these where ranges overlap.
### sizes
```sh
npx node-modules-inspector report sizes --json --limit 20
```
Options:
- `--limit <n>` — cap result count (default: `50`)
- `--include-workspace` — include workspace packages (default: excluded; they have no meaningful install size)
Output shape:
```json
[
{
"spec": "typescript@6.0.3",
"name": "typescript",
"version": "6.0.3",
"workspace": false,
"bytes": 24346827,
"categories": {
"js": { "bytes": 15344521, "count": 200 },
"dts": { "bytes": 7002306, "count": 150 }
}
}
]
```
`categories` keys come from a fixed set: `js`, `ts`, `dts`, `json`, `bin`, `wasm`, `map`, `image`, `css`, `html`, `comp`, `doc`, `test`, `flow`, `other`. Entries are sorted by `bytes` descending.
### maintainers
```sh
npx node-modules-inspector report maintainers --json
```
Options:
- `--sort <depth|migration|latest>` — sort by consumer depth, max migration ratio, or latest release time (default: `depth`)
- `--author <handle>` — filter to consumers maintained by this author; repeatable
- `--no-publint` — exclude publint findings
- `--no-latest-only` — include consumer packages that are not on their latest major
- `--limit <n>` — cap result count
Output shape:
```json
[
{
"consumer": { "spec": "rollup-plugin-esbuild@6.2.1", "name": "rollup-plugin-esbuild", "version": "6.2.1", "depth": 1 },
"authors": [{ "type": "github", "github": "egoist", "avatar": "..." }],
"items": [
{
"kind": "dep-upgrade",
"depName": "unplugin-utils",
"depType": "prod",
"declaredRange": "^0.2.4",
"rawRange": "catalog:deps",
"catalogName": "deps",
"installedHighestVersion": "0.3.1",
"installedHighestSpec": "unplugin-utils@0.3.1",
"installedVersions": ["0.2.4", "0.3.1"],
"migratedCount": 10,
"totalCount": 11,
"migrationRatio": 0.909
},
{
"kind": "publint",
"messages": [/* publint Message objects */],
"counts": { "error": 0, "warning": 1, "suggestion": 2 }
}
],
"maxMigrationRatio": 0.909,
"latestReleasedAt": 1739000000000
}
]
```
How to read this:
- A `dep-upgrade` item means: this consumer declares `depName` at `declaredRange`, but there's a newer installed version (`installedHighestVersion`) that the range does not satisfy. `migrationRatio` is the fraction of consumers in the same cohort that already migrated — a high ratio (e.g. 0.9) means most other consumers already moved on, so this one is lagging.
- `rawRange` differs from `declaredRange` only when the consumer used a pnpm catalog reference (`catalog:deps`); `declaredRange` is the resolved range.
- A `publint` item carries the raw publint messages, partitioned by severity in `counts`.
- `authors` come from the consumer's `package.json` author/maintainers fields, with GitHub-handle detection.
Publint findings only appear when `pkg.resolved.publint` was populated. Enable that by adding `publint: true` to `node-modules-inspector.config.ts` (or by using the project's web UI which runs publint async).
## MCP mode
```sh
npx node-modules-inspector mcp
```
Starts an MCP stdio server. Exposes three tools, identical surface to the CLI:
- `nmi:report-duplicates`
- `nmi:report-sizes`
- `nmi:report-maintainers`
When configured in an MCP client (e.g. Claude Code) under server name `node-modules-inspector`, address them as `node-modules-inspector:nmi:report-duplicates`, etc.
Tool input schemas mirror the CLI options. Tool output is JSON in the exact shape shown above for each report.
Prefer MCP when:
- Multiple queries are expected in one session — the dependency tree is read once and cached across tool calls.
- The agent needs structured output schemas to drive validation.
Prefer the CLI (`report ... --json`) when shell-pipelining (`jq`, redirect, etc.) is more convenient.
## Flags the agent should know
- The first run reads `node_modules` end-to-end and caches npm metadata on disk (under `~/.node-modules-inspector` or similar). Subsequent runs are much faster.
- Workspace packages are excluded from `sizes` by default — pass `--include-workspace` if you actually want them.
- `--depth 8` is enough for almost all real projects. Increase only if the user explicitly asks about deeply-nested transitive dependencies.
- For very large monorepos, running against a single workspace package via `--root packages/<name>` is faster than the whole repo.
## Failure modes
- "No package manager detected" — the project has no `node_modules` directory, or none of pnpm/npm/bun lockfiles. Suggest the user run install first.
- Empty `duplicates` result — fine, it means everything is deduped (mention `pnpm dedupe` etc. only if user wants to verify).
- Empty `maintainers` result — usually means there are no deAsk, share a tip or report that it does not work. Sorted by what people marked helpful.
No posts yet. Be the first to ask or share.
Posts here are written by users, not the PopNx team.