docs-mcp-server

Name: docs-mcp-server
Author: arabold

2.3kinstalls/mo · npm1.6kstars167forks

How to install docs-mcp-server?

arabold/docs-mcp-server

$git clone https://github.com/arabold/docs-mcp-server

Installs into the current project.

Summary

Docs MCP Server solves the problem of AI hallucinations and outdated knowledge by providing a personal, always-current documentation index for your AI coding assistant. It fetches official docs from websites, GitHub, npm, PyPI, and local files, allowing your AI to query the e

Files · 1

View on GitHub

README.md

1	# Grounded Docs: Your AI's Up-to-Date Documentation Expert
2
3	Docs MCP Server solves the problem of AI hallucinations and outdated knowledge by providing a personal, always-current documentation index for your AI coding assistant. It fetches official docs from websites, GitHub, npm, PyPI, and local files, allowing your AI to query the exact version you are using.
4
5	![Docs MCP Server Web Interface](docs/docs-mcp-server.png)
6
7	## ✨ Why Grounded Docs MCP Server?
8
9	The open-source alternative to Context7, Nia, and Ref.Tools.
10
11	- ✅ Up-to-Date Context: Fetches documentation directly from official sources on demand.
12	- 🎯 Version-Specific: Queries target the exact library versions in your project.
13	- 💡 Reduces Hallucinations: Grounds LLMs in real documentation.
14	- 🔒 Private & Local: Runs entirely on your machine; your code never leaves your network.
15	- 🧩 Broad Compatibility: Works with any MCP-compatible client (Claude, Cline, etc.).
16	- 📁 Multiple Sources: Index websites, GitHub repositories, local folders, and zip archives.
17	- 📄 Rich File Support: Processes HTML, Markdown, PDF, Office documents (Word, Excel, PowerPoint), OpenDocument, RTF, EPUB, Jupyter Notebooks, and [90+ source code languages](docs/concepts/supported-formats.md).
18
19	---
20
21	## 📄 Supported Formats
22
23	\| Category \| Formats \|
24	\|----------\|---------\|
25	\| Documents \| PDF, Word (.docx/.doc), Excel (.xlsx/.xls), PowerPoint (.pptx/.ppt), OpenDocument (.odt/.ods/.odp), RTF, EPUB, FictionBook, Jupyter Notebooks \|
26	\| Archives \| ZIP, TAR, gzipped TAR (contents are extracted and processed individually) \|
27	\| Web \| HTML, XHTML \|
28	\| Markup \| Markdown, MDX, reStructuredText, AsciiDoc, Org Mode, Textile, R Markdown \|
29	\| Source Code \| TypeScript, JavaScript, Python, Go, Rust, C/C++, Java, Kotlin, Ruby, PHP, Swift, C#, and [many more](docs/concepts/supported-formats.md#source-code) \|
30	\| Data \| JSON, YAML, TOML, CSV, XML, SQL, GraphQL, Protocol Buffers \|
31	\| Config \| Dockerfile, Makefile, Terraform/HCL, INI, dotenv, Bazel \|
32
33	See [Supported Formats](docs/concepts/supported-formats.md) for the complete reference including MIME types and processing details.
34
35	---
36
37	## 🚀 Quick Start
38
39	### CLI First
40
41	For agents and scripts, the CLI is usually the simplest way to use Grounded Docs.
42
43	1. Index documentation (requires Node.js 22+):
44
45	```bash
46	npx @arabold/docs-mcp-server@latest scrape react https://react.dev/reference/react
47	```
48
49	For hash-routed SPA docs sites, enable hash preservation explicitly:
50
51	```bash
52	npx @arabold/docs-mcp-server@latest scrape my-spa https://docs.example.com/#/guide --preserve-hashes
53	```
54
55	2. Query the index:
56
57	```bash
58	npx @arabold/docs-mcp-server@latest search react "useEffect cleanup" --output yaml
59	```
60
61	3. Fetch a single page as Markdown:
62
63	```bash
64	npx @arabold/docs-mcp-server@latest fetch-url https://react.dev/reference/react/useEffect
65	```
66
67	### Output Behavior
68
69	- Structured commands default to clean JSON on stdout in non-interactive runs.
70	- Use `--output json\|yaml\|toon` to pick a structured format.
71	- Plain-text commands such as `fetch-url` keep their text payload on stdout.
72	- Diagnostics go through the shared logger and are kept off stdout in non-interactive runs.
73	- Use `--quiet` to suppress non-error diagnostics or `--verbose` to enable debug output.
74
75	### Agent Skills
76
77	The [`skills/`](skills/) directory contains [Agent Skills](https://agentskills.io) that teach AI coding assistants how to use the CLI — covering documentation search, index management, and URL fetching.
78
79	### MCP Server
80
81	If you want a long-running MCP endpoint for Claude, Cline, Copilot, Gemini CLI, or other MCP clients:
82
83	1. Start the server:
84
85	```bash
86	npx @arabold/docs-mcp-server@latest
87	```
88
89	2. Open the Web UI at [http://localhost:6280](http://localhost:6280) to add documentation.
90
91	3. Connect your AI client by adding this to your MCP settings (e.g., `claude_desktop_config.json`):
92
93	```json
94	{
95	"mcpServers": {
96	"docs-m

People who liked this also viewed

6 picks

Type

docs-mcp-server

2.3kinstalls/mo · npm1.6kstars167forks

View on GitHub

How to install docs-mcp-server?

arabold/docs-mcp-server

$git clone https://github.com/arabold/docs-mcp-server

Installs into the current project.

Summary

Files · 1

View on GitHub

README.md

1	# Grounded Docs: Your AI's Up-to-Date Documentation Expert
2
3	Docs MCP Server solves the problem of AI hallucinations and outdated knowledge by providing a personal, always-current documentation index for your AI coding assistant. It fetches official docs from websites, GitHub, npm, PyPI, and local files, allowing your AI to query the exact version you are using.
4
5	![Docs MCP Server Web Interface](docs/docs-mcp-server.png)
6
7	## ✨ Why Grounded Docs MCP Server?
8
9	The open-source alternative to Context7, Nia, and Ref.Tools.
10
11	- ✅ Up-to-Date Context: Fetches documentation directly from official sources on demand.
12	- 🎯 Version-Specific: Queries target the exact library versions in your project.
13	- 💡 Reduces Hallucinations: Grounds LLMs in real documentation.
14	- 🔒 Private & Local: Runs entirely on your machine; your code never leaves your network.
15	- 🧩 Broad Compatibility: Works with any MCP-compatible client (Claude, Cline, etc.).
16	- 📁 Multiple Sources: Index websites, GitHub repositories, local folders, and zip archives.
17	- 📄 Rich File Support: Processes HTML, Markdown, PDF, Office documents (Word, Excel, PowerPoint), OpenDocument, RTF, EPUB, Jupyter Notebooks, and [90+ source code languages](docs/concepts/supported-formats.md).
18
19	---
20
21	## 📄 Supported Formats
22
23	\| Category \| Formats \|
24	\|----------\|---------\|
25	\| Documents \| PDF, Word (.docx/.doc), Excel (.xlsx/.xls), PowerPoint (.pptx/.ppt), OpenDocument (.odt/.ods/.odp), RTF, EPUB, FictionBook, Jupyter Notebooks \|
26	\| Archives \| ZIP, TAR, gzipped TAR (contents are extracted and processed individually) \|
27	\| Web \| HTML, XHTML \|
28	\| Markup \| Markdown, MDX, reStructuredText, AsciiDoc, Org Mode, Textile, R Markdown \|
29	\| Source Code \| TypeScript, JavaScript, Python, Go, Rust, C/C++, Java, Kotlin, Ruby, PHP, Swift, C#, and [many more](docs/concepts/supported-formats.md#source-code) \|
30	\| Data \| JSON, YAML, TOML, CSV, XML, SQL, GraphQL, Protocol Buffers \|
31	\| Config \| Dockerfile, Makefile, Terraform/HCL, INI, dotenv, Bazel \|
32
33	See [Supported Formats](docs/concepts/supported-formats.md) for the complete reference including MIME types and processing details.
34
35	---
36
37	## 🚀 Quick Start
38
39	### CLI First
40
41	For agents and scripts, the CLI is usually the simplest way to use Grounded Docs.
42
43	1. Index documentation (requires Node.js 22+):
44
45	```bash
46	npx @arabold/docs-mcp-server@latest scrape react https://react.dev/reference/react
47	```
48
49	For hash-routed SPA docs sites, enable hash preservation explicitly:
50
51	```bash
52	npx @arabold/docs-mcp-server@latest scrape my-spa https://docs.example.com/#/guide --preserve-hashes
53	```
54
55	2. Query the index:
56
57	```bash
58	npx @arabold/docs-mcp-server@latest search react "useEffect cleanup" --output yaml
59	```
60
61	3. Fetch a single page as Markdown:
62
63	```bash
64	npx @arabold/docs-mcp-server@latest fetch-url https://react.dev/reference/react/useEffect
65	```
66
67	### Output Behavior
68
69	- Structured commands default to clean JSON on stdout in non-interactive runs.
70	- Use `--output json\|yaml\|toon` to pick a structured format.
71	- Plain-text commands such as `fetch-url` keep their text payload on stdout.
72	- Diagnostics go through the shared logger and are kept off stdout in non-interactive runs.
73	- Use `--quiet` to suppress non-error diagnostics or `--verbose` to enable debug output.
74
75	### Agent Skills
76
77	The [`skills/`](skills/) directory contains [Agent Skills](https://agentskills.io) that teach AI coding assistants how to use the CLI — covering documentation search, index management, and URL fetching.
78
79	### MCP Server
80
81	If you want a long-running MCP endpoint for Claude, Cline, Copilot, Gemini CLI, or other MCP clients:
82
83	1. Start the server:
84
85	```bash
86	npx @arabold/docs-mcp-server@latest
87	```
88
89	2. Open the Web UI at [http://localhost:6280](http://localhost:6280) to add documentation.
90
91	3. Connect your AI client by adding this to your MCP settings (e.g., `claude_desktop_config.json`):
92
93	```json
94	{
95	"mcpServers": {
96	"docs-m

People who liked this also viewed

6 picks

Type