GroundTruth FAQs

Question 1

How does GroundTruth ensure documentation is up-to-date?

Accepted Answer

GroundTruth fetches live documentation directly from authoritative sources like `llms.txt` files, Jina Reader for JS-rendered pages, and GitHub repositories. This 'at query time' fetching ensures your AI assistant always has access to the latest library versions, APIs, and security practices.

Question 2

What is GroundTruth and how does it help AI coding assistants?

Accepted Answer

GroundTruth is a self-hosted Model Context Protocol (MCP) server that provides AI coding assistants with real-time, accurate documentation and best practices. It prevents the generation of deprecated or insecure code by feeding LLMs the most current information, solving the problem of models hallucinating outdated APIs.

Question 3

What kind of code issues can GroundTruth detect?

Accepted Answer

Its comprehensive `gt_audit` tool scans your source files for over 100 patterns across various categories, including security (XSS, SQL injection, path traversal), performance, accessibility (WCAG AA), React, Next.js, TypeScript, Node.js, and Python. It pinpoints issues at `file:line` with live fix guidance.

Question 4

Does GroundTruth have any usage limits or require API keys?

Accepted Answer

No, a key benefit of GroundTruth is that it runs on your machine as a self-hosted MCP server. This means there are no rate limits, cold starts, or API keys required for its core functionality, offering developers uninterrupted access to real-time context and auditing.

Question 5

Which AI coding environments does GroundTruth integrate with?

Accepted Answer

GroundTruth is designed to integrate seamlessly with popular AI coding assistants and environments that support the Model Context Protocol (MCP), including Claude Code, Cursor, Claude Desktop, and VS Code.

GroundTruth

GroundTruth

主な機能

ユースケース

主な機能

ユースケース