Add configuration management and enhance server functionality with new vector operations

Dockerfile

@@ -7,7 +7,12 @@ RUN npm ci --omit=dev
 
 COPY . .
 
+RUN mkdir -p /data && cp config.json /data/config.json
+
 ENV PORT=7860
+ENV DATA_DIR=/data
 EXPOSE 7860
 
+VOLUME ["/data"]
+
 CMD ["node", "server.js"]

config.json

@@ -0,0 +1,6 @@
+{
+  "CHROMA_URL": "https://maxxcarl-chroma.hf.space",
+  "CHROMA_COLLECTION": "mcptest22",
+  "EMBEDDER_BASE_URL": "https://maxxcarl-emb1024.hf.space",
+  "EMBEDDER_API_KEY": "Azure123"
+}

package-lock.json

@@ -10,6 +10,8 @@
       "license": "ISC",
       "dependencies": {
         "@modelcontextprotocol/sdk": "^1.29.0",
+        "chroma": "^0.0.1",
+        "chromadb": "^3.4.3",
         "express": "^5.2.1"
       }
     },
@@ -173,6 +175,119 @@
         "url": "https://github.com/sponsors/ljharb"
       }
     },
+    "node_modules/chroma": {
+      "version": "0.0.1",
+      "resolved": "https://registry.npmjs.org/chroma/-/chroma-0.0.1.tgz",
+      "integrity": "sha512-spXrvnDrn4IKEN/s0G9vri2+GGZhB15bDU22bbPRnOSEnkGgzGVSFoj/2TeRO5PBysOI4bcHor/8wIamocN09A=="
+    },
+    "node_modules/chromadb": {
+      "version": "3.4.3",
+      "resolved": "https://registry.npmjs.org/chromadb/-/chromadb-3.4.3.tgz",
+      "integrity": "sha512-AouHyTh70Ux1w5TKDmk54wGAszY7NtU/DXyC7o3qGuE8b+b1nT/W3+s8UVd9XwqGZIR4A76+g4u7yJuV/WU3OA==",
+      "license": "Apache-2.0",
+      "dependencies": {
+        "semver": "^7.7.1"
+      },
+      "bin": {
+        "chroma": "dist/cli.mjs"
+      },
+      "engines": {
+        "node": ">=20"
+      },
+      "optionalDependencies": {
+        "chromadb-js-bindings-darwin-arm64": "^1.3.3",
+        "chromadb-js-bindings-darwin-x64": "^1.3.3",
+        "chromadb-js-bindings-linux-arm64-gnu": "^1.3.3",
+        "chromadb-js-bindings-linux-x64-gnu": "^1.3.3",
+        "chromadb-js-bindings-win32-x64-msvc": "^1.3.3"
+      }
+    },
+    "node_modules/chromadb-js-bindings-darwin-arm64": {
+      "version": "1.3.4",
+      "resolved": "https://registry.npmjs.org/chromadb-js-bindings-darwin-arm64/-/chromadb-js-bindings-darwin-arm64-1.3.4.tgz",
+      "integrity": "sha512-pDdWCcR0hCz3pSDiIijBito7YG6FumewfzYE2mIjSwjrO1CKSfQKLwDdTVK4ZNpVGQa16ePoMX+pg7ShSUARJg==",
+      "cpu": [
+        "arm64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": ">= 10"
+      }
+    },
+    "node_modules/chromadb-js-bindings-darwin-x64": {
+      "version": "1.3.4",
+      "resolved": "https://registry.npmjs.org/chromadb-js-bindings-darwin-x64/-/chromadb-js-bindings-darwin-x64-1.3.4.tgz",
+      "integrity": "sha512-5vKVFXSFo+S9qC9by/7oofjcXqQK4IGHiUnPrvTfc7B52gNlPSKeyDO1wha556COc3KtXSZjICEH0nYBJ8UXng==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": ">= 10"
+      }
+    },
+    "node_modules/chromadb-js-bindings-linux-arm64-gnu": {
+      "version": "1.3.4",
+      "resolved": "https://registry.npmjs.org/chromadb-js-bindings-linux-arm64-gnu/-/chromadb-js-bindings-linux-arm64-gnu-1.3.4.tgz",
+      "integrity": "sha512-ELiqDZzU5mZd1BvZugGrhoMkVeNyQaa+PGizd06WIpIJVoHY+S+9ZBjdeM6ZkRnL3vTxD79XBrosxNWhzqy/kg==",
+      "cpu": [
+        "arm64"
+      ],
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">= 10"
+      }
+    },
+    "node_modules/chromadb-js-bindings-linux-x64-gnu": {
+      "version": "1.3.4",
+      "resolved": "https://registry.npmjs.org/chromadb-js-bindings-linux-x64-gnu/-/chromadb-js-bindings-linux-x64-gnu-1.3.4.tgz",
+      "integrity": "sha512-HmTUe7PEIaBngCQ70Yyle81z2wYyg9OMgJYyRUNBYCBp4ED6zTmdaOro41Vs/c/h2UyQeBXFIKNmHDeD2Nr2fw==",
+      "cpu": [
+        "x64"
+      ],
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">= 10"
+      }
+    },
+    "node_modules/chromadb-js-bindings-win32-x64-msvc": {
+      "version": "1.3.4",
+      "resolved": "https://registry.npmjs.org/chromadb-js-bindings-win32-x64-msvc/-/chromadb-js-bindings-win32-x64-msvc-1.3.4.tgz",
+      "integrity": "sha512-punrofQRzKspNMxHmv24qoPeRycV2T3KfedekGV4lOuPPG14i6qggS4x/OWLSAK7ozPfBbrzCejvCa5wfvujhQ==",
+      "cpu": [
+        "x64"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">= 10"
+      }
+    },
     "node_modules/content-disposition": {
       "version": "1.1.0",
       "resolved": "https://registry.npmjs.org/content-disposition/-/content-disposition-1.1.0.tgz",
@@ -903,6 +1018,18 @@
       "integrity": "sha512-YZo3K82SD7Riyi0E1EQPojLz7kpepnSQI9IyPbHHg1XXXevb5dJI7tpyN2ADxGcQbHG7vcyRHk0cbwqcQriUtg==",
       "license": "MIT"
     },
+    "node_modules/semver": {
+      "version": "7.8.1",
+      "resolved": "https://registry.npmjs.org/semver/-/semver-7.8.1.tgz",
+      "integrity": "sha512-rkVq3IXh+4FDGch+KwzX3aV9W3kO54GyEgpvBzSyctDA6Xtd7RJQV1xmXbeQp5v7+VzLOfVqiutSE6GICgPFvg==",
+      "license": "ISC",
+      "bin": {
+        "semver": "bin/semver.js"
+      },
+      "engines": {
+        "node": ">=10"
+      }
+    },
     "node_modules/send": {
       "version": "1.2.1",
       "resolved": "https://registry.npmjs.org/send/-/send-1.2.1.tgz",

package.json

@@ -17,6 +17,8 @@
   "type": "commonjs",
   "dependencies": {
     "@modelcontextprotocol/sdk": "^1.29.0",
+    "chroma": "^0.0.1",
+    "chromadb": "^3.4.3",
     "express": "^5.2.1"
   }
 }

server.js

@@ -1,4 +1,6 @@
 const http = require("node:http");
+const fs = require("node:fs");
+const path = require("node:path");
 const { Server } = require("@modelcontextprotocol/sdk/server/index.js");
 const {
   SSEServerTransport,
@@ -7,27 +9,35 @@ const {
   ListToolsRequestSchema,
   CallToolRequestSchema,
 } = require("@modelcontextprotocol/sdk/types.js");
+const {
+  addDocuments,
+  addDocumentsToolDefinition,
+  listCollectionData,
+  listCollectionDataToolDefinition,
+  queryCollectionData,
+  queryCollectionDataToolDefinition,
+} = require("./vector");
 
 const PORT = Number(process.env.PORT || 3000);
 const HOST = process.env.HOST || "0.0.0.0";
+const DATA_DIR = process.env.DATA_DIR || "/data";
+const CONFIG_PATH = path.join(DATA_DIR, "config.json");
+
+function loadConfig() {
+  try {
+    if (!fs.existsSync(CONFIG_PATH)) {
+      console.error(`Config file not found at ${CONFIG_PATH}`);
+      process.exit(1);
+    }
+    const config = JSON.parse(fs.readFileSync(CONFIG_PATH, "utf8"));
+    return config;
+  } catch (error) {
+    console.error(`Failed to load config from ${CONFIG_PATH}:`, error.message);
+    process.exit(1);
+  }
+}
 
-const sampleData = [
-  {
-    id: "alpha",
-    name: "Alpha record",
-    value: "This is the first sample item.",
-  },
-  {
-    id: "beta",
-    name: "Beta record",
-    value: "This is the second sample item.",
-  },
-  {
-    id: "biggger",
-    name: "coonot picker  record",
-    value: "This is the third sample item.",
-  },
-];
+const config = loadConfig();
 
 function createMcpServer() {
   const server = new Server(
@@ -44,82 +54,49 @@ function createMcpServer() {
 
   server.setRequestHandler(ListToolsRequestSchema, async () => ({
     tools: [
-      {
-        name: "ping",
-        description: "Check that the MCP server is reachable.",
-        inputSchema: {
-          type: "object",
-          properties: {},
-          additionalProperties: false,
-        },
-      },
-      {
-        name: "list_sample_data",
-        description: "Return all sample records for VS Code testing.",
-        inputSchema: {
-          type: "object",
-          properties: {},
-          additionalProperties: false,
-        },
-      },
-      {
-        name: "get_sample_data",
-        description: "Return one sample record by id.",
-        inputSchema: {
-          type: "object",
-          properties: {
-            id: {
-              type: "string",
-              description: "Record id such as alpha, beta, or gamma.",
-            },
-          },
-          required: ["id"],
-          additionalProperties: false,
-        },
-      },
+      addDocumentsToolDefinition,
+      listCollectionDataToolDefinition,
+      queryCollectionDataToolDefinition,
     ],
   }));
 
   server.setRequestHandler(CallToolRequestSchema, async (request) => {
     const { name, arguments: args = {} } = request.params;
 
-    if (name === "ping") {
+    if (name === "list_collection_data") {
+      const results = await listCollectionData(args);
+
       return {
-        content: [{ type: "text", text: "pong" }],
+        content: [
+          {
+            type: "text",
+            text: JSON.stringify(results, null, 2),
+          },
+        ],
       };
     }
 
-    if (name === "list_sample_data") {
+    if (name === "add_documents") {
+      const results = await addDocuments(args);
+
       return {
         content: [
           {
             type: "text",
-            text: JSON.stringify(sampleData, null, 2),
+            text: JSON.stringify(results, null, 2),
           },
         ],
       };
     }
 
-    if (name === "get_sample_data") {
-      const record = sampleData.find((item) => item.id === args.id);
-
-      if (!record) {
-        return {
-          content: [
-            {
-              type: "text",
-              text: `No sample record found for id "${args.id}".`,
-            },
-          ],
-          isError: true,
-        };
-      }
+    if (name === "query_collection_data") {
+      const results = await queryCollectionData(args);
 
       return {
         content: [
           {
             type: "text",
-            text: JSON.stringify(record, null, 2),
+            text: JSON.stringify(results, null, 2),
           },
         ],
       };
@@ -179,7 +156,7 @@ async function main() {
     const url = new URL(req.url, `http://${hostHeader}`);
 
     if (req.method === "GET" && url.pathname === "/health") {
-      sendJson(res, 200, { status: "nigg" });
+      sendJson(res, 200, { status: "ok" });
       return;
     }
 

vector.js

@@ -0,0 +1,320 @@
+const fs = require("node:fs");
+const path = require("node:path");
+
+async function fetchJson(url, options = {}) {
+  const response = await fetch(url, options);
+  const text = await response.text();
+  let payload;
+
+  try {
+    payload = text ? JSON.parse(text) : undefined;
+  } catch (error) {
+    throw new Error(`Invalid JSON response from ${url}`);
+  }
+
+  if (!response.ok) {
+    const message =
+      payload?.error?.message ||
+      payload?.message ||
+      `Request failed with status ${response.status}`;
+    const error = new Error(message);
+    error.status = response.status;
+    error.payload = payload ?? text;
+    throw error;
+  }
+
+  return payload;
+}
+
+function loadConfig() {
+  const configPath = path.join(__dirname, "config.json");
+  const raw = fs.readFileSync(configPath, "utf8");
+  return JSON.parse(raw);
+}
+
+function extractEmbedding(payload) {
+  if (Array.isArray(payload?.embedding)) {
+    return payload.embedding;
+  }
+  if (Array.isArray(payload?.embeddings)) {
+    return payload.embeddings[0];
+  }
+  if (
+    Array.isArray(payload?.data) &&
+    Array.isArray(payload.data[0]?.embedding)
+  ) {
+    return payload.data[0].embedding;
+  }
+  if (Array.isArray(payload)) {
+    return payload[0];
+  }
+  return undefined;
+}
+
+async function getEmbedding(config, input, isQuery = false) {
+  const payload = await fetchJson(`${config.EMBEDDER_BASE_URL}/embed`, {
+    method: "POST",
+    headers: {
+      accept: "application/json",
+      "x-api-key": config.EMBEDDER_API_KEY,
+      "Content-Type": "application/json",
+    },
+    body: JSON.stringify({ input: [input], is_query: isQuery }),
+  });
+
+  const embedding = extractEmbedding(payload);
+  if (!Array.isArray(embedding)) {
+    throw new Error("Embedding response did not include a vector");
+  }
+
+  return embedding;
+}
+
+function buildApiUrl(config, apiBase, pathSegment) {
+  const base = config.CHROMA_URL.replace(/\/+$/, "");
+  const api = apiBase.startsWith("/") ? apiBase : `/${apiBase}`;
+  const path = pathSegment.startsWith("/") ? pathSegment : `/${pathSegment}`;
+  return `${base}${api}${path}`;
+}
+
+async function fetchChroma(config, pathSegment, options = {}) {
+  const url = buildApiUrl(config, "/api/v2", pathSegment);
+  return fetchJson(url, options);
+}
+
+async function ensureDatabase(config) {
+  const tenant = config.CHROMA_TENANT || "default";
+  const database = config.CHROMA_DATABASE || "default";
+
+  try {
+    await fetchChroma(config, `/tenants/${tenant}/databases/${database}`);
+  } catch (error) {
+    if (error.status === 404) {
+      await fetchChroma(config, `/tenants/${tenant}/databases`, {
+        method: "POST",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({ name: database }),
+      });
+    } else {
+      throw error;
+    }
+  }
+}
+
+async function getOrCreateCollection(config) {
+  const tenant = config.CHROMA_TENANT || "default";
+  const database = config.CHROMA_DATABASE || "default";
+
+  const collectionPath = `/tenants/${tenant}/databases/${database}/collections?name=${encodeURIComponent(config.CHROMA_COLLECTION)}`;
+  const existing = await fetchChroma(config, collectionPath);
+  const collection = Array.isArray(existing) ? existing[0] : existing;
+
+  if (collection?.id) {
+    return collection.id;
+  }
+
+  const createPath = `/tenants/${tenant}/databases/${database}/collections`;
+  const created = await fetchChroma(config, createPath, {
+    method: "POST",
+    headers: { "Content-Type": "application/json" },
+    body: JSON.stringify({ name: config.CHROMA_COLLECTION }),
+  });
+
+  return created.id;
+}
+
+async function getCollectionId(config) {
+  const tenant = config.CHROMA_TENANT || "default";
+  const database = config.CHROMA_DATABASE || "default";
+
+  const collectionPath = `/tenants/${tenant}/databases/${database}/collections?name=${encodeURIComponent(config.CHROMA_COLLECTION)}`;
+  const existing = await fetchChroma(config, collectionPath);
+  const collection = Array.isArray(existing) ? existing[0] : existing;
+
+  if (!collection?.id) {
+    throw new Error("Collection not found");
+  }
+
+  return collection.id;
+}
+
+const addDocumentsToolDefinition = {
+  name: "add_documents",
+  description: "Add documents into the Chroma collection using the embedder.",
+  inputSchema: {
+    type: "object",
+    properties: {
+      documents: {
+        type: "array",
+        description: "Documents to add to the collection.",
+        items: {
+          type: "object",
+          properties: {
+            id: { type: "string", description: "Unique document id." },
+            text: { type: "string", description: "Document text." },
+            metadata: {
+              type: "object",
+              description: "Optional metadata for the document.",
+            },
+          },
+          required: ["id", "text"],
+          additionalProperties: false,
+        },
+        minItems: 1,
+      },
+    },
+    required: ["documents"],
+    additionalProperties: false,
+  },
+};
+
+const listCollectionDataToolDefinition = {
+  name: "list_collection_data",
+  description: "Return all records stored in the Chroma collection.",
+  inputSchema: {
+    type: "object",
+    properties: {
+      limit: {
+        type: "integer",
+        description: "Maximum number of records to return.",
+        minimum: 1,
+      },
+      offset: {
+        type: "integer",
+        description: "Offset for pagination.",
+        minimum: 0,
+      },
+      includeEmbeddings: {
+        type: "boolean",
+        description: "Include embedding vectors in the response.",
+      },
+    },
+    additionalProperties: false,
+  },
+};
+
+const queryCollectionDataToolDefinition = {
+  name: "query_collection_data",
+  description: "Query documents in the Chroma collection using the embedder.",
+  inputSchema: {
+    type: "object",
+    properties: {
+      query: {
+        type: "string",
+        description: "Query text for semantic search.",
+      },
+      nResults: {
+        type: "integer",
+        description: "Number of results to return (default 5).",
+        minimum: 1,
+      },
+      includeEmbeddings: {
+        type: "boolean",
+        description: "Include embedding vectors in the response.",
+      },
+    },
+    required: ["query"],
+    additionalProperties: false,
+  },
+};
+
+async function addDocuments(args = {}) {
+  if (!Array.isArray(args.documents) || args.documents.length === 0) {
+    throw new Error("documents must be a non-empty array");
+  }
+
+  const config = loadConfig();
+  await ensureDatabase(config);
+  const collectionId = await getOrCreateCollection(config);
+  const tenant = config.CHROMA_TENANT || "default";
+  const database = config.CHROMA_DATABASE || "default";
+  const embeddings = [];
+
+  for (const doc of args.documents) {
+    embeddings.push(await getEmbedding(config, doc.text, false));
+  }
+
+  await fetchChroma(
+    config,
+    `/tenants/${tenant}/databases/${database}/collections/${collectionId}/add`,
+    {
+      method: "POST",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({
+        ids: args.documents.map((doc) => doc.id),
+        embeddings,
+        documents: args.documents.map((doc) => doc.text),
+        metadatas: args.documents.map((doc) => doc.metadata ?? null),
+      }),
+    },
+  );
+
+  return { added: args.documents.length, documents: args.documents };
+}
+
+async function listCollectionData(args = {}) {
+  const config = loadConfig();
+  await ensureDatabase(config);
+  const collectionId = await getOrCreateCollection(config);
+  const tenant = config.CHROMA_TENANT || "default";
+  const database = config.CHROMA_DATABASE || "default";
+  const include = ["documents", "metadatas"];
+
+  if (args.includeEmbeddings) {
+    include.push("embeddings");
+  }
+
+  const body = { include };
+  if (Number.isInteger(args.limit)) {
+    body.limit = args.limit;
+  }
+  if (Number.isInteger(args.offset)) {
+    body.offset = args.offset;
+  }
+
+  return fetchChroma(
+    config,
+    `/tenants/${tenant}/databases/${database}/collections/${collectionId}/get`,
+    {
+      method: "POST",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify(body),
+    },
+  );
+}
+
+async function queryCollectionData(args = {}) {
+  const config = loadConfig();
+  const tenant = config.CHROMA_TENANT || "default";
+  const database = config.CHROMA_DATABASE || "default";
+  const collectionId = await getCollectionId(config);
+  const include = ["documents", "metadatas", "distances"];
+
+  if (args.includeEmbeddings) {
+    include.push("embeddings");
+  }
+
+  const queryEmbedding = await getEmbedding(config, args.query, true);
+  return fetchChroma(
+    config,
+    `/tenants/${tenant}/databases/${database}/collections/${collectionId}/query`,
+    {
+      method: "POST",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({
+        query_embeddings: [queryEmbedding],
+        n_results: args.nResults ?? 5,
+        include,
+      }),
+    },
+  );
+}
+
+module.exports = {
+  addDocuments,
+  addDocumentsToolDefinition,
+  listCollectionData,
+  listCollectionDataToolDefinition,
+  queryCollectionData,
+  queryCollectionDataToolDefinition,
+};