Claude
Claude Desktop and Claude Code support MCP natively. Add an entry to your MCP config
(.mcp.json or Claude Desktop's settings) with this server's URL, then restart the client.
MCP
MCP
The Model Context Protocol (MCP) is an open standard that lets AI assistants call tools over a simple JSON-RPC interface instead of guessing from training data. This server exposes Lucee's own function and tag reference, full-text doc search, and CFML AST parsing as MCP tools — so an assistant can look up exact signatures, search recipes, and inspect real CFML source before it answers.
Single JSON-RPC 2.0 endpoint — no separate path per tool. Point your MCP client's server URL at the address below.
| Tool | Arguments | Description |
|---|---|---|
get_lucee_function |
name | FLD descriptor for a built-in Lucee function. |
get_lucee_tag |
name | TLD descriptor for a Lucee tag. |
search_lucee_docs |
query, maxResults | Full-text search across functions, tags, and recipes. |
parse_cfml_ast |
source or path, mode, summary, maxDepth | Parse CFML source into an AST tree or compact summary. |
query_cfml_ast |
source or path, nodeType, name, line | Find matching nodes in already-parsed CFML. |
Add the endpoint above as an MCP server in your client of choice:
Claude Desktop and Claude Code support MCP natively. Add an entry to your MCP config
(.mcp.json or Claude Desktop's settings) with this server's URL, then restart the client.
Add the endpoint as a custom connector under Settings → Connectors (Developer Mode), so ChatGPT can call the Lucee tools during a conversation.
Register the endpoint as an MCP server in the Gemini CLI's settings (or your Gemini client's MCP configuration) to give Gemini the same tool access.