English · ç®€ä½“ä¸æ–‡ · 日本語 · Español · Français
A curated marketplace of developer tools for Claude Code.
No bloat. Tested. Works in any AI editor.
One install command. Every plugin wires up its MCP tools, slash commands, and agents automatically.
claude plugin marketplace add Aboudjem/10xTip
One install command. Every plugin wires up its MCP tools, slash commands, and agents automatically.
Six plugins today. Every one of them passes the quality bar below.
Tip
Want to see them run? Each tool repo has its own demo: sniff, ui-ux-suite, recap-studio, aws-cost-audit, goalify, and humanizer.
AI-powered QA that walks your running app like a user, discovers its flows, and finds real bugs across accessibility, performance, and broken interactions. Auto-detects your dev server. A source scan is also available via sniff scan. No API keys, no config.
claude plugin install sniff@10x # as a plugin (primary)
npx sniff-qa # as a standalone CLI
# one-line multi-CLI install (Cursor, Codex, Gemini, Copilot, and more):
curl -fsSL https://raw.githubusercontent.com/Aboudjem/sniff/main/install.sh | bash -s codex|
Slash commands
|
MCP tools
|
GitHub →  · npm →
Scans your CSS, JSX, and Tailwind config. Scores 12 design dimensions: accessibility, color system, typography, layout and spacing, component quality, visual hierarchy, interaction quality, responsiveness, visual polish, performance UX, information architecture, and platform appropriateness. Cites the UX law each finding violates, then shows you the exact fix. It audits, it never edits. Zero dependencies, 100% local.
claude plugin install ui-ux-suite@10x # as a plugin (primary)
npx ui-ux-suite # as a standalone CLI
# one-line multi-CLI install (Cursor, Codex, Gemini, Copilot, and more):
curl -fsSL https://raw.githubusercontent.com/Aboudjem/ui-ux-suite/main/install.sh | bash -s codex|
Slash commands
|
MCP tools
|
GitHub →  · npm →
Turn any topic or coding session into a clean, mobile-first one-page explainer a smart 18-year-old can read in 5 minutes. Output is a self-contained, offline HTML page. Deterministic heuristic checks validate clarity, accessibility, and structure before you ship, and every claim is cited.
claude plugin install recap-studio@10x # as a plugin (primary)
/recap "Latest AI models" # in any Claude Code session
# one-line multi-CLI install (Cursor, Codex, Gemini, Copilot, and more):
curl -fsSL https://raw.githubusercontent.com/Aboudjem/recap-studio/main/install.sh | bash -s codex|
Slash commands
|
What you get
|
GitHub →  · Architecture →
Reads your live AWS account (read-only by default), attributes every dollar, and finds idle, orphaned, and over-provisioned resources. Verifies every price live against the AWS Price List API (it hardcodes none) and gives each finding a confidence level, the evidence behind it, and a rollback. Nothing gets deleted without proof and your sign-off.
claude plugin install aws-cost-audit@10x # as a plugin (primary)
# one-line install into another CLI's skills directory:
curl -fsSL https://raw.githubusercontent.com/Aboudjem/aws-cost-audit-skill/main/install.sh | bash -s codex|
What it does
|
Safe by default
|
Scopes a big coding task while it still has your context: it reads and researches the repo, locks the few real decisions, then writes a self-contained, self-deleting /goal file with the finish line defined as commands the run can check. You /clear and run that file; a fresh full-context session executes the whole job, tests as it goes, and deletes the file only when every success criterion passes. A Claude Code skill (no MCP server, no CLI).
claude plugin install goalify@10x # as a plugin (primary)
/goalify <your task> # then, in any Claude Code session|
What it does
|
Safe by default
|
Detects 43 AI writing patterns (inflated symbolism, promotional language, em dash overuse, rule-of-three, AI vocabulary words, negative parallelisms, vague attributions) and rewrites the text with sentence-length burstiness so it stops reading like a model wrote it. Pick from 5 voice profiles, get a 0-100 AI-tell score on demand, and run it in detect, rewrite, or edit mode. Pure Markdown, zero dependencies, no network calls. A Claude Code skill (no MCP server, no CLI).
claude plugin install humanizer@10x # as a plugin (primary)
/humanizer "<your text>" # then, in any Claude Code session|
What it does
|
Three modes
|
Every 10x tool plugin is dual mode: install it as a Claude Code plugin, or run it as a plain MCP server in the editor of your choice. (aws-cost-audit, goalify, and humanizer are Claude Code skills rather than MCP servers, so they run in Claude Code: aws-cost-audit reads your AWS account locally, goalify writes a /goal file you run in a fresh session, and humanizer rewrites the text you pass it.)
npx sniff-qa --mcp
npx ui-ux-suite --mcp
# recap-studio also ships an MCP server (see its README; requires a local build)Each tool plugin also ships a one-line install.sh that wires it into another CLI's skills directory in a single command (curl -fsSL https://raw.githubusercontent.com/Aboudjem/<repo>/main/install.sh | bash -s <cli>), plus .cursor-plugin and .copilot-plugin manifests. Each project's README has copy-paste snippets for Cursor, VS Code + Copilot, Codex, Gemini, Windsurf, and Continue.dev.
Every plugin here passes this bar. If it stops passing, it gets removed.
| Zero bloat | Vanilla Node.js, no runtime dependencies |
| One-command install | No config files, no API keys |
| Real tests | Not aspirational, not "coming soon" |
| Dual mode | Works as a Claude Code plugin AND as an MCP server |
| Actively maintained | Shipped this quarter, not abandoned last year |
| No telemetry | Runs locally, your code never leaves your machine |
The full eight-point checklist (32 items) lives in QUALITY-BAR.md. Maintainers re-verify every plugin against it on a quarterly review.
Note
Current roster: sniff 0.7.0 (441 tests), ui-ux-suite 0.5.0 (311 tests), recap-studio 0.4.0 (43 tests across 5 test-bearing packages), aws-cost-audit 0.2.0 (built test-first, independently verified end-to-end on a live account, zero hardcoded prices), goalify 1.0.0 (built test-first; manifest + skill-lint tests and a recorded RED→GREEN eval baseline on Haiku, Sonnet, and Opus), humanizer 0.1.0 (pure Markdown, zero dependencies, no network calls; detects 43 AI writing patterns with a 0-100 AI-tell score). Last verified 2026-05-30.
Got a plugin that belongs here? See CONTRIBUTING.md.
If 10x helped you ship better code, consider starring it.
It helps other devs find these tools.
Built by Adam Boudjemaa · MIT License · No telemetry · No data collection
