feat: add 2 tools for agents (execute SPARQL query and search examples/classes extracted from endpoints)

Author: vemonet

CONTRIBUTING.md

@@ -48,8 +48,8 @@ npm version patch
 
 To do:
 
-- [ ] Include a MCP server with tools: getExamples, getClassesSchema, use the [`ai-embed-search`](https://www.npmjs.com/package/ai-embed-search) package? Or LanceDB? [simile](https://github.com/iaavas/simile-search)?
-- [ ] Make the fetch request sent for executing the SPARQL query configurable at the endpoint level: GET/POST, headers, basic auth
+- [ ] When we store the class schema should we store it as a formatted string, so we can also easily support extracting and using SHACL/ShEx shapes without having to parse them
+- [ ] Make the fetch request sent for executing the SPARQL query configurable at the endpoint level: GET/POST, headers, basic auth. It can all be configured in a new section inside the backend expand
 - [ ] Enable to have backend profiles with a set of pre-optimized completion queries for a given triplestore/ontology.
 - [ ] Make it so format put all comments `#+` at the start of the query (in qlue-ls)
 

README.md

@@ -7,6 +7,7 @@ A Visual Studio Code extension with support for the [SPARQL](https://www.w3.org/
 - **Syntax highlighting** using semantic tokens to adapt to themes
 - **Query execution & results inspection** with the [YASGUI](https://github.com/rdfjs/Yasgui) YASR component
 - **Configuration** of the language server
+- **Tools** for agents: execute query, and search in examples/classes schema extracted from the endpoints
 
 ## 📥 Installation