# anthropics/claude-plugins-community

Source: [GitHub](https://github.com/anthropics/claude-plugins-community)  
Feed7 permalink: https://feed7.dev/p/claude-plugins-community-0vhi4d8  
Published: Unknown  
Trust: Needs Review (needs_review)

## Why Included

Anthropic now exposes its reviewed community plugin catalog as a nightly read-only mirror for Claude Code and Cowork. Installation is direct, but submissions and changes must pass its separate review pipeline.

## Source Summary

This repository mirrors the community plugin marketplace for **Claude Code and Claude Cowork**. Its marketplace file syncs **nightly** from Anthropic's internal pipeline, and every listed plugin has passed automated security scanning plus distribution approval.

## Practical Implication

Builders can add the marketplace from the Claude CLI and install a named community plugin, while using the repository as an inspectable catalog. Plugin authors must submit through Anthropic's directory form rather than opening a pull request.

## Agent-Ready Context

This repository mirrors the community plugin marketplace for **Claude Code and Claude Cowork**. Its marketplace file syncs **nightly** from Anthropic's internal pipeline, and every listed plugin has passed automated security scanning plus distribution approval.

Builders can add the marketplace from the Claude CLI and install a named community plugin, while using the repository as an inspectable catalog. Plugin authors must submit through Anthropic's directory form rather than opening a pull request.

The repository is **read-only**, so it is not the source of review decisions or direct community edits. Automated scanning and approval reduce one class of risk, but the material does not describe review criteria, ongoing monitoring, or guarantees about plugin behavior.

## Connected Context

Feed7 judgment across 551 accumulated Signals:

This establishes an official, inspectable distribution layer for Claude plugins rather than another discovery list. Nightly synchronization, scanning, and approval add governance before installation, but the read-only mirror centralizes control and narrows the safety claim to undisclosed pre-distribution checks rather than ongoing behavioral assurance.

- [VoltAgent/awesome-agent-skills](https://feed7.dev/p/awesome-agent-skills-0vfknie) — The two catalogs serve discovery differently: this mirror adds Claude-specific scanning and approval, while VoltAgent offers wider cross-assistant coverage without common security or compatibility guarantees.
- [anthropics/claude-code](https://feed7.dev/p/claude-code-0pmb63d) — Claude Code is the installation and execution surface for this marketplace, so its plugin support enables adoption while its deployment and privacy considerations remain relevant.
- [cursor/plugins](https://feed7.dev/p/plugins-0vuoqp2) — Cursor provides a comparable manifest-based plugin distribution model, but the Claude mirror adds centralized scanning and approval without revealing whether packaged behavior receives ongoing monitoring.
- [alirezarezvani/claude-skills](https://feed7.dev/p/claude-skills-06ix38u) — That cross-agent catalog can distribute through plugin marketplaces and includes its own auditor, showing that marketplace approval and repository-level review are complementary rather than interchangeable controls.

## Context Map

- Layer: tools
- Domains: coding
- Topics: skills, tool-use, coding-agents

## Uncertainty

- The repository is **read-only**, so it is not the source of review decisions or direct community edits. Automated scanning and approval reduce one class of risk, but the material does not describe review criteria, ongoing monitoring, or guarantees about plugin behavior.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.