@vespermcp/mcp-server 1.2.21 → 1.2.22

package/build/tools/formatter.js

@@ -54,7 +54,7 @@ export function formatSearchResults(results) {
     results.forEach((ds, index) => {
         const relevanceScore = ds.relevance_score || 0;
         // Source badge and access level
-        const openSources = ["huggingface", "uci", "github", "worldbank", "nasa"];
+        const openSources = ["huggingface", "openml", "s3", "uci", "github", "worldbank", "nasa"];
         const isOpen = openSources.includes(ds.source);
         const sourceLabel = ds.source.charAt(0).toUpperCase() + ds.source.slice(1);
         const accessBadge = isOpen ? "Open Access" : "Requires API Key";
@@ -131,7 +131,7 @@ export function formatDatasetInfo(ds) {
     output += `${ds.name}\n`;
     output += "═".repeat(80) + "\n\n";
     // Source and safety
-    const openSources = ["huggingface", "uci", "github", "worldbank", "nasa"];
+    const openSources = ["huggingface", "openml", "s3", "uci", "github", "worldbank", "nasa"];
     const isOpen = openSources.includes(ds.source);
     const sourceLabel = ds.source.charAt(0).toUpperCase() + ds.source.slice(1);
     const accessBadge = isOpen ? "Open Access" : "Requires API Key";
@@ -149,7 +149,10 @@ export function formatDatasetInfo(ds) {
     output += `Safety: ${safetyIndicator}\n`;
     output += `ID: ${ds.id}\n\n`;
     if (!isOpen && ds.source === "kaggle") {
-        output += `NOTE: This dataset requires a Kaggle API key (KAGGE_USERNAME/KAGGLE_KEY) to prepare.\n\n`;
+        output += `NOTE: This dataset uses the Kaggle connector. Vesper can access it through server-managed credentials when configured, otherwise a Kaggle key is still required.\n\n`;
+    }
+    if (!isOpen && ds.source === "dataworld") {
+        output += `NOTE: This dataset uses the data.world connector. Vesper can access it through a server-managed token when configured.\n\n`;
     }
     // Description
     if (ds.description) {

package/build/utils/python-runtime.js

@@ -0,0 +1,130 @@
+import { spawn } from "child_process";
+import fs from "fs";
+import os from "os";
+import path from "path";
+function getHomeDir(buildDir) {
+    return os.homedir() || process.env.HOME || process.env.USERPROFILE || buildDir;
+}
+export function getVesperDataRoot(buildDir = process.cwd()) {
+    return path.join(getHomeDir(buildDir), ".vesper");
+}
+export function getManagedPythonPath(buildDir = process.cwd()) {
+    const dataRoot = getVesperDataRoot(buildDir);
+    return process.platform === "win32"
+        ? path.join(dataRoot, ".venv", "Scripts", "python.exe")
+        : path.join(dataRoot, ".venv", "bin", "python");
+}
+function getFallbackPythonCommand() {
+    return process.platform === "win32" ? "py" : "python3";
+}
+export function resolvePythonCommand(buildDir = process.cwd()) {
+    const managedPython = getManagedPythonPath(buildDir);
+    if (fs.existsSync(managedPython)) {
+        return managedPython;
+    }
+    const envPython = process.env.VESPER_PYTHON;
+    if (envPython) {
+        return envPython;
+    }
+    const localCandidates = process.platform === "win32"
+        ? [
+            path.resolve(buildDir, ".venv", "Scripts", "python.exe"),
+            path.resolve(buildDir, "..", ".venv", "Scripts", "python.exe")
+        ]
+        : [
+            path.resolve(buildDir, ".venv", "bin", "python"),
+            path.resolve(buildDir, "..", ".venv", "bin", "python")
+        ];
+    for (const candidate of localCandidates) {
+        if (fs.existsSync(candidate)) {
+            return candidate;
+        }
+    }
+    return getFallbackPythonCommand();
+}
+function runPythonCommand(pythonPath, args, timeoutMs = 300000) {
+    return new Promise((resolve, reject) => {
+        const proc = spawn(pythonPath, args, {
+            env: {
+                ...process.env,
+                PYTHONIOENCODING: "utf-8",
+            },
+        });
+        let stdout = "";
+        let stderr = "";
+        const timer = setTimeout(() => {
+            proc.kill();
+            resolve({ code: 124, stdout, stderr: stderr || `Python command timed out after ${timeoutMs}ms` });
+        }, timeoutMs);
+        proc.stdout.on("data", (data) => {
+            stdout += data.toString();
+        });
+        proc.stderr.on("data", (data) => {
+            stderr += data.toString();
+        });
+        proc.on("close", (code) => {
+            clearTimeout(timer);
+            resolve({ code: code ?? 1, stdout, stderr });
+        });
+        proc.on("error", (error) => {
+            clearTimeout(timer);
+            reject(error);
+        });
+    });
+}
+async function createManagedPythonEnv(buildDir) {
+    const dataRoot = getVesperDataRoot(buildDir);
+    const venvDir = path.join(dataRoot, ".venv");
+    const managedPython = getManagedPythonPath(buildDir);
+    if (fs.existsSync(managedPython)) {
+        return managedPython;
+    }
+    fs.mkdirSync(dataRoot, { recursive: true });
+    const bootstrapAttempts = process.platform === "win32"
+        ? [
+            { command: "py", args: ["-3", "-m", "venv", venvDir] },
+            { command: "python", args: ["-m", "venv", venvDir] },
+        ]
+        : [
+            { command: "python3", args: ["-m", "venv", venvDir] },
+            { command: "python", args: ["-m", "venv", venvDir] },
+        ];
+    let lastError = "";
+    for (const attempt of bootstrapAttempts) {
+        try {
+            const result = await runPythonCommand(attempt.command, attempt.args, 180000);
+            if (result.code === 0 && fs.existsSync(managedPython)) {
+                await runPythonCommand(managedPython, ["-m", "pip", "install", "--disable-pip-version-check", "--upgrade", "pip"], 300000);
+                return managedPython;
+            }
+            lastError = (result.stderr || result.stdout || "Unknown venv creation error").trim();
+        }
+        catch (error) {
+            lastError = error?.message || String(error);
+        }
+    }
+    throw new Error(`Failed to create Vesper Python environment. ${lastError}`.trim());
+}
+export async function ensurePythonPackages(buildDir, requirements) {
+    const pythonPath = await createManagedPythonEnv(buildDir).catch(() => resolvePythonCommand(buildDir));
+    const missing = [];
+    for (const requirement of requirements) {
+        const check = await runPythonCommand(pythonPath, [
+            "-c",
+            `import importlib.util,sys; sys.exit(0 if importlib.util.find_spec(${JSON.stringify(requirement.module)}) else 1)`
+        ], 20000);
+        if (check.code !== 0) {
+            missing.push(requirement);
+        }
+    }
+    if (missing.length === 0) {
+        return pythonPath;
+    }
+    const packages = [...new Set(missing.map(requirement => requirement.packageName))];
+    const install = await runPythonCommand(pythonPath, ["-m", "pip", "install", "--disable-pip-version-check", ...packages], 600000);
+    if (install.code !== 0) {
+        const details = (install.stderr || install.stdout || "Unknown pip install error").trim();
+        throw new Error(`Failed to install Python packages (${packages.join(", ")}). ${details}`);
+    }
+    return pythonPath;
+}

package/package.json

@@ -1,17 +1,16 @@
 {
   "name": "@vespermcp/mcp-server",
-  "version": "1.2.21",
+  "version": "1.2.22",
   "description": "AI-powered dataset discovery, quality analysis, and preparation MCP server with multimodal support (text, image, audio, video)",
   "type": "module",
   "main": "build/index.js",
   "bin": {
-    "mcp-server": "./build/index.js",
-    "vespermcp": "./build/index.js",
-    "vesper-wizard": "scripts/wizard.js"
+    "vesper-wizard": "wizard.cjs"
   },
   "files": [
     "build/**/*",
     "src/python/**/*",
+    "wizard.cjs",
     "scripts/**/*",
     "README.md",
     "LICENSE",
@@ -31,6 +30,7 @@
     "fuse": "node build/index.js fuse",
     "discover": "node build/index.js discover",
     "download": "node build/index.js download",
+    "export": "node build/index.js export",
     "config": "node build/index.js config",
     "test-fusion-engine": "py src/python/test_fusion_engine.py",
     "setup": "node build/index.js --setup",
@@ -59,7 +59,7 @@
   "license": "MIT",
   "repository": {
     "type": "git",
-    "url": "https://github.com/vesper/mcp-server"
+    "url": "git+https://github.com/vesper/mcp-server.git"
   },
   "engines": {
     "node": ">=18.0.0",
@@ -68,6 +68,8 @@
   "dependencies": {
     "@huggingface/hub": "^2.7.1",
     "@modelcontextprotocol/sdk": "^1.25.2",
+    "@polar-sh/nextjs": "^0.9.4",
+    "@supabase/supabase-js": "^2.98.0",
     "@xenova/transformers": "^2.17.2",
     "adm-zip": "^0.5.16",
     "ajv": "^8.17.1",

package/scripts/postinstall.cjs

@@ -2,13 +2,34 @@
 
 const { execSync } = require('child_process');
 const fs = require('fs');
+const os = require('os');
 const path = require('path');
 
 console.log('\n🚀 Setting up Vesper MCP Server...\n');
 
+function getPythonBootstrapCommand() {
+    const attempts = process.platform === 'win32'
+        ? ['py -3', 'python']
+        : ['python3', 'python'];
+
+    for (const command of attempts) {
+        try {
+            execSync(`${command} --version`, { stdio: 'pipe' });
+            return command;
+        } catch {
+            // try next command
+        }
+    }
+
+    return null;
+}
+
 // 1. Check for Python
+const pythonBootstrap = getPythonBootstrapCommand();
 try {
-    execSync('python --version', { stdio: 'pipe' });
+    if (!pythonBootstrap) {
+        throw new Error('Python not found');
+    }
     console.log('✅ Python found');
 } catch (e) {
     console.warn('⚠️  Python not found. Please install Python 3.8+ for full functionality.');
@@ -16,36 +37,15 @@ try {
     process.exit(0); // Don't fail installation
 }
 
-// 2. Install Python dependencies
-console.log('\n📦 Installing Python dependencies...');
-const pythonPackages = [
-    'opencv-python',
-    'pillow',
-    'numpy',
-    'librosa',
-    'soundfile',
-    'aiohttp',
-    'aiofiles',
-    'datasets',
-    'webdataset',
-    'kaggle'
-];
-
-try {
-    execSync(`python -m pip install ${pythonPackages.join(' ')}`, {
-        stdio: 'inherit',
-        timeout: 120000 // 2 minutes timeout
-    });
-    console.log('✅ Python dependencies installed');
-} catch (e) {
-    console.warn('⚠️  Failed to install some Python dependencies.');
-    console.warn('   You may need to install them manually:');
-    console.warn(`   pip install ${pythonPackages.join(' ')}\n`);
-}
-
-// 3. Create data directories
-const homeDir = process.env.HOME || process.env.USERPROFILE;
+const homeDir = os.homedir() || process.env.HOME || process.env.USERPROFILE;
 const vesperDataDir = path.join(homeDir, '.vesper');
+const managedVenvDir = path.join(vesperDataDir, '.venv');
+const managedPython = process.platform === 'win32'
+    ? path.join(managedVenvDir, 'Scripts', 'python.exe')
+    : path.join(managedVenvDir, 'bin', 'python');
+const requirementsPath = path.resolve(__dirname, '..', 'src', 'python', 'requirements.txt');
+
+// 2. Create data directories
 const dirs = [
     vesperDataDir,
     path.join(vesperDataDir, 'data'),
@@ -62,7 +62,63 @@ dirs.forEach(dir => {
 
 console.log(`✅ Data directories created at ${vesperDataDir}`);
 
-// 4. Auto-configure Claude Desktop (Best Effort)
+// 3. Create a managed Vesper Python environment
+console.log('\n🐍 Preparing managed Python environment...');
+try {
+    if (!fs.existsSync(managedPython)) {
+        execSync(`${pythonBootstrap} -m venv "${managedVenvDir}"`, {
+            stdio: 'inherit',
+            timeout: 180000,
+        });
+    }
+    console.log(`✅ Managed Python ready at ${managedVenvDir}`);
+} catch (e) {
+    console.warn('⚠️  Failed to create the managed Vesper Python environment.');
+    console.warn(`   Vesper will fall back to PATH Python and may need to self-heal at runtime. ${(e && e.message) || ''}`.trim());
+}
+
+// 4. Install Python dependencies into the managed environment
+console.log('\n📦 Installing Python dependencies...');
+const pythonPackages = [
+    'opencv-python',
+    'pillow',
+    'librosa',
+    'soundfile',
+    'pyarrow'
+];
+
+try {
+    const targetPython = fs.existsSync(managedPython) ? `"${managedPython}"` : pythonBootstrap;
+    execSync(`${targetPython} -m pip install --disable-pip-version-check --upgrade pip`, {
+        stdio: 'inherit',
+        timeout: 180000,
+    });
+    execSync(`${targetPython} -m pip install --disable-pip-version-check -r "${requirementsPath}" ${pythonPackages.join(' ')}`, {
+        stdio: 'inherit',
+        timeout: 600000,
+    });
+    console.log('✅ Python dependencies installed');
+} catch (e) {
+    console.warn('⚠️  Failed to install some Python dependencies.');
+    console.warn('   You may need to install them manually into the Vesper runtime:');
+    console.warn(`   ${fs.existsSync(managedPython) ? managedPython : pythonBootstrap} -m pip install -r "${requirementsPath}" ${pythonPackages.join(' ')}\n`);
+}
+
+// 5. Rebuild better-sqlite3 for current Node.js version
+console.log('\n🔧 Rebuilding native modules for current Node.js...');
+try {
+    execSync('npm rebuild better-sqlite3', {
+        stdio: 'pipe',
+        timeout: 60000,
+        cwd: path.resolve(__dirname, '..')
+    });
+    console.log('✅ Native modules rebuilt successfully');
+} catch (e) {
+    console.warn('⚠️  Could not rebuild better-sqlite3: ' + (e.message || e));
+    console.warn('   If you see ERR_DLOPEN_FAILED, run: npm rebuild better-sqlite3');
+}
+
+// 6. Auto-configure Claude Desktop (Best Effort)
 console.log('\n⚙️  Attempting to auto-configure Claude Desktop...');
 
 function getClaudeConfigPath() {