Skip to skills
AI Craft Pad
Home/Skills/MCP Servers/Context7
MCP Servers

Context7

An MCP documentation service with declared remote and local package options, suited to questions tied to a specific library and version.

upstash/context7 ↗License: MITRuntime safety: UnknownScore: 75/100

What it does

The registry snapshot lists a hosted HTTP endpoint plus npm and MCP bundle distributions. Repository documentation describes resolving a library identity and retrieving relevant documentation. This can add source context to a coding task, but it does not prove that generated code uses the returned API correctly or that every library version is covered.

Best for

A developer who can name the intended library, installed version and narrow API question. Supply those details before asking for an implementation, then compare the returned reference with the dependency already used by the project.

Tradeoffs

Hosted requests involve an external service and its current terms. The registry declares optional secret references; the README recommends authentication for higher limits. CLI setup and MCP setup are different paths, and their runtime requirements should not be assumed identical.

A practical first task

Choose either the remote connection or a declared local distribution using the publisher's client-specific guide. Ask one version-sensitive question, record the matched library identifier and keep the relevant source reference with your implementation notes.

What to validate

Check the library match, target version, relevance of returned material and behavior when documentation is unavailable. Treat retrieved pages as untrusted reference data. AI Craft Pad read metadata and documentation without calling the service or testing its accuracy.

Compatibility

cursor, opencode, claude-code

Repository setup documentation; protocol compatibility is not a runtime test

Compatibility source

Installation and requirements

For the declared remote endpoint, use an HTTP-capable MCP client and the publisher authentication flow if needed. Local npm/MCP-bundle distributions have separate runtime requirements.

Remote endpoint: https://mcp.context7.com/mcp. The registry declares an optional Authorization secret reference. Follow the source client guide and store any credential through your client; do not place keys in shared configuration or URLs.

Use your agent’s MCP settings and the publisher’s exact setup. Match the declared transport and package version. Credentials belong in your client’s supported secret storage. No command on this page is run by AI Craft Pad.

npm package

@upstash/context7-mcp · version 4.4.4 · stdio · runtime Unknown

CONTEXT7_API_KEY — optional, secret reference

mcpb package

https://github.com/upstash/context7/releases/download/mcpb-v4.4.4/context7.mcpb · version 4.4.4 · stdio · runtime Unknown

CONTEXT7_API_KEY — optional, secret reference

streamable-http endpoint

https://mcp.context7.com/mcp

Declared registry endpoint; availability, authentication, service cost and runtime behavior have not been tested.

Authorization — optional, secret reference

Read publisher setup ↗

Unknown; source software and connected services may have separate terms.

Safety and permissions

Unknown — no runtime security assessment.

Only sanitized manifest/registry metadata and bounded documentation are inspected. Package code, transitive dependencies, hooks and remote services have not been audited. Review access before use.

  • credential-reference
  • network-access
  • shell-or-package-install

Related Skills

Category-based discovery suggestions. Review functional overlap; these combinations have not been integration-tested.

AI & machine learning87/100

huggingface local models

huggingface/skills

A workflow for selecting GGUF model files and running or serving them locally with llama.cpp.

First-party sourceApache-2.0Editorial review

Repository star change: +2 over 26.7 hours. 2026-11-01

11,243 repo stars2 text flagsReview ↗
AI & machine learning87/100

transformers js

huggingface/skills

A JavaScript inference reference for model pipelines, WebGPU or WASM execution, quantization and resource management.

First-party sourceApache-2.0Editorial review

Repository star change: +2 over 26.7 hours. 2026-11-01

11,243 repo stars2 text flagsReview ↗
AI & machine learning84/100

runway dev models

runwayml/skills

Integrate an exact Runway model endpoint behind an application server and handle its task lifecycle.

First-party sourceMITEditorial review
74 repo stars2 text flagsReview ↗

Related Plugins

Category-based discovery suggestions. Review functional overlap; these combinations have not been integration-tested.

AI & machine learning75/100

learning output style

anthropics/claude-plugins-official · Plugin

Explore learning-output-style as an agent plugin package for ai & machine learning. Review its supported format, bundled components and setup requirements before installing.

Manifest metadataApache-2.0
45,633 repo stars1 metadata flagReview ↗
AI & machine learning75/100

ralph loop

anthropics/claude-plugins-official · Plugin

Explore ralph-loop as an agent plugin package for ai & machine learning. Review its supported format, bundled components and setup requirements before installing.

Manifest metadataApache-2.0
45,633 repo stars1 metadata flagReview ↗
AI & machine learning75/100

c level skills

alirezarezvani/claude-skills · Plugin

Explore c-level-skills as an agent plugin package for ai & machine learning. Review its supported format, bundled components and setup requirements before installing.

Manifest metadataMIT
29,099 repo stars1 metadata flagReview ↗

Similar tools and alternatives

Category-based discovery suggestions. Review functional overlap; these combinations have not been integration-tested.

AI & machine learning72/100

Klarion

0x1Adi/Klarion · MCP server

Explore Klarion as an MCP connection for ai & machine learning. Compare the declared transport, package requirements and source evidence before connecting an agent.

Registry metadataMIT
11 repo stars2 metadata flagsReview ↗
AI & machine learning72/100

AI Proof of Us

0xddneto/AI-Proof-of-Us · MCP server

Explore AI Proof of Us as an MCP connection for ai & machine learning. Compare the declared transport, package requirements and source evidence before connecting an agent.

Registry metadataMIT
4 repo stars2 metadata flagsReview ↗
AI & machine learning71/100

Silk

21J3phy/silk · MCP server

Explore Silk as an MCP connection for ai & machine learning. Compare the declared transport, package requirements and source evidence before connecting an agent.

Registry metadataApache-2.0
1 repo stars1 metadata flagReview ↗

Works well with

Choose a complementary capability only after checking for duplicated permissions and bundled connections. Agent Stacks explain the roles and validation steps; they are proposals, not tested integrations.

Build a stack for your task · Related AI guides · Report a correction