tei-mcp
Helps AI agents read and write valid TEI XML. Element lookup, attribute resolution, content model expansion, nesting validation, document validation, and ODD customisation.
README
tei-mcp
<p align="center"> <img src="banner.png" alt="tei-mcp — TEI P5 for AI agents" width="600"> </p>
<p align="center"> <a href="https://doi.org/10.5281/zenodo.19039570"><img src="https://img.shields.io/badge/DOI-10.5281%2Fzenodo.19039570-blue" alt="DOI"></a> <a href="https://pypi.org/project/tei-mcp/"><img src="https://img.shields.io/pypi/v/tei-mcp" alt="PyPI version"></a> <a href="https://github.com/Pantagrueliste/tei-mcp/blob/master/LICENSE"><img src="https://img.shields.io/pypi/l/tei-mcp" alt="License"></a> <a href="https://modelcontextprotocol.io"><img src="https://img.shields.io/badge/MCP-compatible-green" alt="MCP compatible"></a> <a href="https://tei-c.org/guidelines/"><img src="https://img.shields.io/badge/TEI-P5-orange" alt="TEI P5"></a> </p>
An MCP server that helps AI agents read and write valid TEI XML. It parses the TEI P5 specification and exposes 16 tools for element lookup, attribute resolution, content model expansion, nesting validation, document validation, and ODD customisation.
Features
- Element, class, macro, and module lookup with case-insensitive matching and typo suggestions
- Attribute resolution across the full TEI class hierarchy (local + inherited)
- Content model expansion into structured trees with class and macro resolution
- Nesting validation (direct parent-child and recursive reachability with path tracking)
- Document validation against TEI P5: content models, attributes, closed value lists, reference integrity, deprecation warnings
- Single-element validation for incremental editing workflows
- ODD customisation support: load a project ODD to constrain the schema (moduleRef filtering, elementSpec delete/change, attDef modifications)
- Regex search across all entity types (elements, classes, macros, modules)
- Deprecation awareness with validUntil dates and replacement suggestions
- Attribute suggestion by intent description (keyword matching against attribute descriptions)
Requirements
- Python 3.10+
- uv (recommended) or pip
Installation
The quickest way is via uvx, which fetches and runs the server automatically:
uvx tei-mcp
Or install from PyPI:
pip install tei-mcp
Or clone and install from source:
git clone https://github.com/Pantagrueliste/tei-mcp.git
cd tei-mcp
uv sync
On first run, the server downloads p5subset.xml from the TEI website (~5 MB) and caches it locally.
Usage
tei-mcp works with any MCP-compatible client. Add the following to your client's MCP server configuration:
{
"mcpServers": {
"tei": {
"command": "uvx",
"args": ["tei-mcp"]
}
}
}
Where this file lives depends on your client:
| Client | Configuration file |
|---|---|
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) |
| Claude Code | .mcp.json in your project directory |
| Cursor | .cursor/mcp.json in your project directory |
| Windsurf | ~/.codeium/windsurf/mcp_config.json |
| Other clients | Consult your client's MCP documentation |
Standalone
uvx tei-mcp
The server communicates over stdio using the MCP protocol.
Tools
| Tool | Description |
|---|---|
lookup_element |
Look up an element by name (e.g., persName) |
lookup_class |
Look up a class by name (e.g., att.global) |
lookup_macro |
Look up a macro by name (e.g., macro.paraContent) |
list_module_elements |
List all elements in a module (e.g., namesdates) |
search |
Regex search across all TEI entities |
list_attributes |
Resolve all attributes for an element (local + inherited) |
class_membership_chain |
Show the full class hierarchy chain |
expand_content_model |
Expand content model into a structured tree |
valid_children |
List all valid direct children of an element |
check_nesting |
Check if an element can appear inside another |
check_nesting_batch |
Check multiple nesting pairs in one call |
suggest_attribute |
Find relevant attributes by intent description |
validate_document |
Validate a TEI XML file against the spec |
validate_element |
Validate a single element in context |
load_customisation |
Load an ODD customisation file |
unload_customisation |
Clear the loaded customisation |
Most tools accept use_odd=True to query the customised schema instead of the full TEI P5.
ODD Customisation
Load a project-specific ODD file to constrain the schema:
1. Call load_customisation("/path/to/my-project.odd")
2. Use use_odd=True on subsequent tool calls
3. Call unload_customisation() to revert to the full spec
Supported ODD features:
moduleRefwithinclude/exceptfilteringelementSpec mode="delete"to remove elementselementSpec mode="change"withattDefmodifications (delete, change, add)- Closed/semi value list restrictions
Environment Variables
| Variable | Default | Description |
|---|---|---|
TEI_ODD_PATH |
— | Path to a local p5subset.xml (skips download) |
TEI_ODD_URL |
TEI-C GitHub URL | Custom URL for the ODD file |
Development
# Install dev dependencies
uv sync
# Run tests
uv run pytest
# Run tests with coverage info
uv run pytest -v
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.