Is RuleSell affiliated with Anthropic?

No. RuleSell has zero affiliation with Anthropic, Anysphere (Cursor), or any AI vendor. It's a two-person, bootstrapped marketplace. The official Claude plugin marketplace is run by Anthropic itself — that first-party status is its biggest advantage, and we say so on this page.

Does the official Claude plugin marketplace verify plugins?

Partially. It runs a two-tier review: all submissions get a baseline check, and a subset that passes stricter human review earns an 'Anthropic Verified' badge. But the repo README states plainly that 'Anthropic does not control what MCP servers, files, or other software are included in plugins and cannot verify that they will work as intended.' Verified means reviewed, not guaranteed — treat unverified plugins like any code from the internet.

Does RuleSell security-scan listings?

No automated security scanning is in production, and we won't pretend otherwise. What runs today: a regex gate on uploads that rejects content matching known-bad patterns (live API keys, eval() calls, shell-injection shapes), plus manual admin review for flagged listings. Semgrep rules and malware scanning are on the roadmap. The /trust page is the live record of which trust signals are real and which are planned.

What does RuleSell's quality grade actually measure?

Three computed signals, with public weights: freshness (0.30 — step decay on the last push date), schema cleanliness (0.35 — valid category, a real SPDX license, a description of at least 80 characters, preview content of at least 200), and review score (0.35 — mean rating × 20 once a listing has 3+ reviews, neutral 60 before that). 85+ is an A, 70+ a B, 50+ a C. GitHub stars are not an input. Two more signals — token efficiency and install success — are designed but not shipped.

Is RuleSell free? Is the official marketplace free?

Both are free today. The official marketplace is free by design and pays creators nothing. RuleSell is free during beta; paid publishing with creator payouts is the stated post-beta path (Stripe Connect is the planned rail), and no take rate has been committed publicly yet. If you want a free catalog forever, the official marketplace is the safer bet. If you're a creator who wants the option to charge, only one of these two has that on its roadmap.

Should I use the official marketplace or RuleSell?

Both, for different jobs. If you live in Claude Code and want a plugin installed in one command with first-party review behind it, use the official marketplace — it wins that job outright. Use RuleSell when you need configs that work across Cursor, Windsurf, Cline, or Codex too, when you want a quality signal before installing (stars don't provide one), or when you want to find the original author behind a listing.

Claude Plugin Marketplace vs RuleSell (June 2026): Honest Decision Guide

# Claude Plugin Marketplace vs RuleSell: Honest Decision Guide (June 2026) Anthropic shipped its official Claude Code plugin marketplace on May 22, 2026. The repo sat at 29.7k stars by June 9, install is one command inside the CLI, everything is free, and a subset of plugins carries an "Anthropic Verified" badge. If your whole AI-tooling life is Claude Code, that marketplace is the default — and this page will not talk you out of it. RuleSell is a different thing wearing a similar coat: a 315-listing catalog that spans Claude Code, Cursor, Windsurf, Cline, and Codex, puts a computed A/B/C quality grade on every entry, attributes each open-source listing to its real GitHub author (who can claim it via OAuth), and plans paid publishing for creators after beta. One of us wrote this page, so read it with that in mind — every claim below carries a source or a number you can check.

Who wins at what

The table above is the answer. First-party trust, in-CLI install, catalog momentum, and being free with no commercial layer all go to the official marketplace. Cross-tool coverage, a quality grade per listing, author attribution with a claim flow, and a paid path for creators all go to RuleSell. Four wins each is not diplomacy — it's what falls out when you score each dimension separately.

Where the official marketplace wins decisively

First-party trust. It's Anthropic's own directory for Anthropic's own tool. Submissions go through a two-tier review, and plugins that pass the stricter human tier get the "Anthropic Verified" badge. No third party can match that signal for Claude Code plugins, including us. Zero install friction. Plugins install from inside Claude Code itself. No browser, no copy-paste, no account. For a Claude Code user, the distance between "found it" and "running it" is one command. RuleSell's flow — find on the web, copy the install path or use the CLI — has more steps, full stop. Catalog momentum. 29.7k stars and 3.2k forks as of June 9, 2026, up from 28.7k ten days earlier. The ecosystem submits through a public form and the catalog compounds with the community's attention. A two-person marketplace does not out-momentum that, and we won't claim to. Free, with no commercial layer. There is no pricing page because there is no pricing. If a commercial layer anywhere in the stack bothers you, the official marketplace removes the question.

Where RuleSell wins decisively

Coverage beyond Claude Code. The official marketplace is Claude Code plugins only — that's its job. RuleSell's listings carry environment metadata across Claude Code, Cursor, Windsurf, Cline, Codex, and other targets, so "will this MCP server also work in Cursor?" is answered on the listing instead of in a GitHub issue. If your team runs more than one tool, a single-tool directory answers half your question. A quality grade you can check. Every RuleSell listing gets a 0–100 score from three computed signals — freshness (0.30), schema cleanliness (0.35), review score (0.35) — mapped to A/B/C grades. The weights are public and GitHub stars are deliberately not an input, because stars measure past popularity, not current maintenance: our catalog has A-grade MCP servers from sub-100-star repos pushed in the last three weeks, and C-grade listings from multi-thousand-star repos whose last push was 2023. The official marketplace's signal is binary (Verified or not); a graded signal tells you more about the long tail. Author attribution and the claim flow. Each open-source listing shows the original author, SPDX license, star count, and last-push date, and the author can claim the listing via GitHub OAuth to edit or remove it. The official marketplace lists what's submitted; RuleSell's catalog includes indexed OSS work, and the attribution-plus-claim mechanic is how that stays honest rather than scraped-and-rebranded. A paid path for creators. The official marketplace pays creators nothing, by design. RuleSell is free during beta, and paid publishing with creator payouts is the stated post-beta path. The take rate isn't committed yet — we'd rather say that than invent a number — but the direction is the structural difference: one of these platforms intends to route money to config authors.

Where the comparison gets uncomfortable

"Verified" is doing less work than it sounds like — on both sides. The official repo's own README says Anthropic "does not control what MCP servers, files, or other software are included in plugins and cannot verify that they will work as intended." And RuleSell runs no automated security scanning either: today's protection is a regex gate on uploads (live API keys, eval(), shell-injection patterns) plus manual review when something gets flagged. Semgrep and malware scanning are roadmap items, not features. Neither platform should be your security boundary — review what you install. SentinelOne's research on dependency hijack in Claude Code marketplaces is worth reading before you install anything from anywhere, including us. The official marketplace's existence helps RuleSell's category — which cuts both ways. The directory space went from one registry in December 2025 to eight-plus marketplaces by Q2 2026, after the official launch. That growth validates the demand and crowds the field at the same time. Most of the new entrants are scraped aggregators competing on raw count; a 315-listing catalog only justifies itself if the grading and attribution earn their keep. RuleSell's review signal is thin during beta. Most listings don't yet have the 3+ reviews needed for a non-neutral review score, so today's grade differentiation comes mostly from freshness and schema cleanliness. That's stated on the /trust page and it's stated here. The grade is honest about what it measures; it doesn't yet measure everything we want it to.

When to use both

The realistic setup for a multi-tool team:

Official marketplace for Claude Code plugins — in-CLI install, Verified badge where available, free.

RuleSell for cross-tool questions — "does this exist for Cursor and Cline?", "which of these five similar MCP servers is actually maintained?", "who wrote this and is it claimed?"

Your own review before running anything from either, because neither platform verifies that code is safe.

If you only use Claude Code and never look at quality beyond the Verified badge, you may not need RuleSell at all. That's an honest possible outcome of reading this page.

Where this comparison fails / what we don't know

RuleSell wrote this comparison about itself. We've anchored every claim to a number, a public repo, or a page you can load — but the dimension selection is still ours, and a dimension list is where bias hides. We also don't publish traffic or user counts for RuleSell here because we don't have audited numbers worth citing, and we won't decorate the page with soft ones. We don't know how the official marketplace evolves. Anthropic could add cross-tool support, quality grades, or payouts, and any of those would erase a row of the table above. The catalog-momentum gap also means the official marketplace's network effects compound faster than ours — the honest read is that RuleSell's lane only exists as long as it does the things the official one doesn't.