scai 0.1.28 → 0.1.30

package/dist/commands/AskCmd.js

@@ -1,5 +1,7 @@
 import readline from 'readline';
-import { searchFiles } from "../db/fileIndex.js";
+import path from 'path';
+import { searchFiles, queryFiles } from "../db/fileIndex.js";
+import { sanitizeQueryForFts } from "../utils/sanitizeQuery.js";
 import { generate } from "../lib/generate.js";
 export async function runAskCommand(query) {
     if (!query) {
@@ -12,21 +14,43 @@ export async function runAskCommand(query) {
     }
     console.log(`🔍 Searching for: "${query}"\n`);
     const start = Date.now();
-    const results = await searchFiles(query, 5);
+    const semanticResults = await searchFiles(query, 5);
     const duration = Date.now() - start;
-    console.log(`⏱️ searchFiles took ${duration}ms and returned ${results.length} result(s)`);
-    if (results.length > 0) {
-        console.log('📊 Closest files based on semantic similarity:');
-        results.forEach((file, i) => {
-            console.log(`  ${i + 1}. 📄 Path: ${file?.path} | Score: ${file?.score?.toFixed(3)}`);
+    console.log(`⏱️ searchFiles took ${duration}ms and returned ${semanticResults.length} result(s)`);
+    // Also run fallback keyword search
+    const safeQuery = sanitizeQueryForFts(query);
+    const fallbackResults = queryFiles(safeQuery, 10);
+    // Merge semantic and fallback results
+    const seen = new Set();
+    const combinedResults = [];
+    for (const file of semanticResults) {
+        const resolved = path.resolve(file.path);
+        seen.add(resolved);
+        combinedResults.push(file); // Already scored
+    }
+    for (const file of fallbackResults) {
+        const resolved = path.resolve(file.path);
+        if (!seen.has(resolved)) {
+            seen.add(resolved);
+            combinedResults.push({
+                path: file.path,
+                summary: file.summary,
+                score: 0.0, // fallback score
+            });
+        }
+    }
+    if (combinedResults.length > 0) {
+        console.log('📊 Closest files based on semantic + keyword match:');
+        combinedResults.forEach((file, i) => {
+            console.log(`  ${i + 1}. 📄 Path: ${file.path} | Score: ${file.score?.toFixed(3) ?? 'fallback'}`);
         });
     }
     else {
-        console.log('⚠️ No similar embeddings found. Asking the model for context instead...');
+        console.log('⚠️ No similar files found. Asking the model for context only...');
     }
-    // 🧠 Use stored summaries directly
+    // Aggregate summaries
     let allSummaries = '';
-    for (const file of results) {
+    for (const file of combinedResults) {
         if (!file?.summary) {
             console.warn(`⚠️ No summary available for file: ${file?.path}`);
             continue;

package/dist/commands/BackupCmd.js

@@ -0,0 +1,12 @@
+// src/commands/BackupCmd.ts
+import { backupScaiFolder } from '../db/backup.js';
+export async function runBackupCommand() {
+    console.log('🔁 Backing up .scai folder...');
+    const result = await backupScaiFolder();
+    if (result) {
+        console.log('✅ Backup complete.');
+    }
+    else {
+        console.error('❌ Backup failed.');
+    }
+}

package/dist/commands/ResetDbCmd.js

@@ -1,34 +1,9 @@
 import fs from 'fs';
-import fsp from 'fs/promises';
-import { db } from '../db/client.js';
-import { DB_PATH, SCAI_HOME } from '../constants.js';
-import lockfile from 'proper-lockfile';
 import path from 'path';
-function getBackupDir() {
-    const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
-    return path.join(SCAI_HOME, `backup-${timestamp}`);
-}
-async function backupScaiFolder() {
-    const backupDir = getBackupDir();
-    try {
-        await fsp.mkdir(backupDir, { recursive: true });
-        const files = await fsp.readdir(SCAI_HOME);
-        for (const file of files) {
-            const srcPath = path.join(SCAI_HOME, file);
-            const destPath = path.join(backupDir, file);
-            const stat = await fsp.stat(srcPath);
-            if (stat.isFile()) {
-                await fsp.copyFile(srcPath, destPath);
-            }
-        }
-        console.log(`📦 Backed up .scai folder to: ${backupDir}`);
-        return backupDir;
-    }
-    catch (err) {
-        console.warn('⚠️ Failed to back up .scai folder:', err instanceof Error ? err.message : err);
-        return null;
-    }
-}
+import lockfile from 'proper-lockfile';
+import { db } from '../db/client.js';
+import { DB_PATH } from '../constants.js';
+import { backupScaiFolder } from '../db/backup.js'; // <-- New import
 export async function resetDatabase() {
     console.log('🔁 Backing up existing .scai folder...');
     await backupScaiFolder();

package/dist/config/IgnoredExtensions.js

@@ -1,68 +1,65 @@
 export const IGNORED_EXTENSIONS = [
-    // Media
+    // ─── Media Files ─────────────────────────────────────────────
     '.png', '.jpg', '.jpeg', '.gif', '.webp', '.svg', '.ico',
     '.mp4', '.mp3', '.mov', '.avi', '.mkv', '.flv', '.wav', '.flac',
     '.aac', '.m4a', '.wma', '.3gp', '.webm', '.ogg', '.aiff', '.au',
-    // Archives & install packages
+    // ─── Archive & Install Packages ─────────────────────────────
     '.zip', '.tar', '.gz', '.bz2', '.xz', '.rar', '.7z',
     '.jar', '.war', '.ear', '.deb', '.rpm', '.pkg', '.msi', '.dmg', '.cab', '.apk',
     '.tar.gz', '.tar.bz2', '.tar.xz', '.tar.lzma', '.tar.zst',
-    // Binaries & executables
+    // ─── Binaries & Executables ─────────────────────────────────
     '.exe', '.dll', '.bin', '.so', '.dylib', '.a', '.lib',
-    '.iso', '.img', '.elf', '.o', '.obj', '.msm', '.vbs', '.jscript',
+    '.iso', '.img', '.elf', '.o', '.obj', '.msm',
     '.cmd', '.bat', '.ps1', '.sh', '.bash', '.run',
-    // Runtime / build / cache
-    '.log', '.tmp', '.map',
+    '.vbs', '.jscript',
+    // ─── Build / Runtime Cache ─────────────────────────────────
+    '.log', '.tmp', '.map', '.tsbuildinfo', '.coverage', '.eslintcache',
     '.db', '.sqlite', '.pkl', '.sav', '.rdb', '.ldb',
-    '.pyc', '.class', '.tsbuildinfo', '.coverage', '.eslintcache',
-    '.yarn', '.webpack', '.babel', '.compilercache',
-    // Fonts & styles
-    '.woff', '.woff2', '.ttf', '.eot', '.otf', '.css.map',
-    '.scss', '.sass', '.less', '.styl',
-    // Certs, keys, credentials
+    '.pyc', '.class', '.yarn', '.webpack', '.babel', '.compilercache',
+    // ─── Fonts & Stylesheets ───────────────────────────────────
+    '.woff', '.woff2', '.ttf', '.eot', '.otf',
+    '.scss', '.sass', '.less', '.styl', '.css.map',
+    // ─── Certificates & Keys ───────────────────────────────────
     '.crt', '.key', '.pem', '.pub', '.asc', '.gpg', '.p12', '.csr', '.der', '.pfx',
-    // Backups / temp
-    '.bak', '.old', '.swp', '.swo', '.orig',
-    '.sublime-workspace', '.sublime-project', '.db-shm', '.db-wal',
-    // System/config folders (still ignored by path, not extension)
-    '.DS_Store', '.bundle', '.npmrc',
-    // GIS / Geospatial
-    '.shp', '.shx', '.dbf', '.prj', '.qix', '.sbn', '.sbx', '.shp.xml', '.cpg', '.gpkg', '.mif', '.mid',
-    // Enterprise BI / Reporting
+    // ─── Backups & Temp Files ──────────────────────────────────
+    '.bak', '.bakx', '.old', '.swp', '.swo', '.orig', '.~',
+    '.db-shm', '.db-wal',
+    '.log.old', '.log.1', '.log.2', '.log.3',
+    '.xml.bck', '.wbk', '.asd',
+    // ─── Editor / IDE Artifacts ────────────────────────────────
+    '.sublime-workspace', '.sublime-project',
+    '.iml', '.ipr', '.iws', '.idea', '.code-workspace', '.vscode',
+    '.ncb', '.suo', '.user', '.sdf',
+    '.classlist', '.nb-gradle', '.gradle',
+    // ─── System & OS Metadata ──────────────────────────────────
+    '.DS_Store', '.ds_store', '.thumb', '.ini', '.sys', '.lnk',
+    '.trash', '.trashes', '.localized', '.spotlight-v100',
+    '.~lock', '.appledouble',
+    // ─── GIS / Geospatial ──────────────────────────────────────
+    '.shp', '.shx', '.dbf', '.prj', '.qix', '.sbn', '.sbx',
+    '.shp.xml', '.cpg', '.gpkg', '.mif', '.mid',
+    // ─── Business Intelligence / Reporting ─────────────────────
     '.pbix', '.rdl', '.rpt', '.bqy', '.iqy',
-    // ETL / DWH / Modeling
+    // ─── ETL / DWH / Modeling ──────────────────────────────────
     '.abf', '.dtsx', '.bim', '.xmi',
-    // CAD / Engineering
+    // ─── CAD / Engineering ─────────────────────────────────────
     '.dwg', '.dxf', '.step', '.stp', '.sldprt', '.sldasm',
     '.iges', '.igs', '.3ds', '.fbx',
-    // Forms / Print / Publishing
+    // ─── Forms / Print / Publishing ────────────────────────────
     '.xps', '.afpub', '.pub', '.indd', '.qxd', '.frm', '.frx', '.frl',
-    // ERP / Finance / Legacy DB
+    // ─── ERP / Finance / Legacy DB ─────────────────────────────
     '.mbd', '.fdb', '.nav', '.accdb', '.mdb', '.gdb',
     '.sap', '.sappkg', '.qbw', '.qbb',
-    // Lock files (but NOT lock *configs*)
-    '.lck', '.lockfile', '.db-lock', '.pid', '.socket',
-    // Documentation and Readme Files
-    '.md', // Markdown files for documentation
-    '.rst', // reStructuredText files for documentation
-    '.txt', // Plain text files (can be used for documentation or notes)
-    // Configuration files (JSON/YAML/XML)
-    '.json', // General JSON config files
-    '.yaml', // YAML configuration files
-    '.yml', // YAML configuration files
-    '.xml', // XML configuration files
-    // Miscellaneous Important Files
-    '.properties', // Java properties files (configuration)
-    '.xml.bck', // Backup XML files
-    '.bak', // General backup files
-    '.swp', // Swap files (e.g., vim/sublime temp files)
-    '.swo', // Swap files
-    '.orig', // Original backup files
-    '.old', // Old version files
-    '.~', // Temp files generated by some editors (e.g., vim)
-    '.md5', // MD5 checksums
-    '.sha1', // SHA1 checksums
-    '.sha256', // SHA256 checksums
-    '.tmp', // Temporary files
-    '.bak', // Backup file (for various applications)
+    // ─── Lock Files (Exclude lock *data*, not lock *configs*) ──
+    '.lck', '.lock', '.lockfile', '.db-lock', '.pid', '.sock', '.socket',
+    // ─── Documentation & Readme ────────────────────────────────
+    '.md', '.rst', '.txt',
+    // ─── Configuration Files ───────────────────────────────────
+    '.json', '.yaml', '.yml', '.xml', '.properties',
+    // ─── Scientific & Data Serialization ───────────────────────
+    '.npy', '.npz', '.feather', '.parquet', '.orc', '.avro',
+    '.hdf5', '.h5', '.rda', '.rds',
+    '.csv.gz', '.dump',
+    // ─── Misc Hashes & Metadata ────────────────────────────────
+    '.md5', '.sha1', '.sha256', '.xlsb', '.xar',
 ];

package/dist/config/StopWords.js

@@ -1,17 +1,15 @@
-/**
- * These common words are ignored from search queries
- * to reduce noise and improve FTS and embedding match quality.
- */
 export const STOP_WORDS = new Set([
     // Articles & conjunctions
-    'a', 'an', 'and', 'but', 'or', 'nor',
+    'a', 'an', 'the', 'and', 'but', 'or', 'nor', 'yet', 'so',
     // Prepositions
-    'at', 'by', 'for', 'from', 'in', 'into', 'of', 'on', 'to', 'with', 'about', 'above', 'below', 'under', 'over', 'through',
+    'at', 'by', 'for', 'from', 'in', 'into', 'of', 'on', 'to', 'with', 'about', 'above',
+    'below', 'under', 'over', 'through', 'within', 'without', 'across', 'behind', 'near',
     // Pronouns
-    'i', 'me', 'my', 'mine', 'you', 'your', 'yours', 'he', 'him', 'his', 'she', 'her', 'hers', 'it', 'its',
-    'we', 'us', 'our', 'ours', 'they', 'them', 'their', 'theirs',
-    // Determiners
+    'i', 'me', 'my', 'mine', 'you', 'your', 'yours', 'he', 'him', 'his', 'she', 'her', 'hers',
+    'it', 'its', 'we', 'us', 'our', 'ours', 'they', 'them', 'their', 'theirs', 'someone', 'something',
+    // Determiners & quantifiers
     'this', 'that', 'these', 'those', 'some', 'any', 'each', 'every', 'either', 'neither',
+    'few', 'several', 'all', 'many', 'much', 'most', 'more', 'less', 'another',
     // Auxiliary and modal verbs
     'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being',
     'do', 'does', 'did',
@@ -19,10 +17,18 @@ export const STOP_WORDS = new Set([
     'can', 'could', 'shall', 'should', 'will', 'would', 'may', 'might', 'must',
     // Wh-words and generic question words
     'what', 'which', 'who', 'whom', 'whose', 'where', 'when', 'why', 'how',
-    // Misc common functional words
-    'as', 'if', 'than', 'then', 'there', 'because', 'so', 'just', 'only', 'not', 'no',
-    // Very common verbs to strip from noisy queries
-    'use', 'get', 'make', 'need', 'want', 'let', 'help', 'work', 'see', 'look', 'like', 'know',
-    // Other frequent noise
-    'all', 'more', 'most', 'many', 'much', 'such', 'also', 'again'
+    // Common verbs (noisy or generic)
+    'use', 'get', 'set', 'make', 'need', 'want', 'let', 'help', 'work',
+    'see', 'look', 'like', 'know', 'tell', 'show', 'call', 'create', 'put', 'run',
+    'start', 'stop', 'begin', 'end', 'open', 'close',
+    // Functional & filler words
+    'as', 'if', 'than', 'then', 'there', 'here', 'because', 'so', 'just', 'only',
+    'not', 'no', 'also', 'again', 'really', 'actually', 'maybe', 'please',
+    'very', 'too', 'quite', 'rather', 'even', 'still', 'ever', 'never',
+    'though', 'although', 'indeed', 'perhaps', 'such',
+    // Command fluff
+    'explain', 'describe', 'define', 'find', 'give', 'provide', 'show',
+    'tell', 'check',
+    // Miscellaneous low-information words
+    'yes', 'no', 'okay', 'ok', 'thing', 'stuff', 'way', 'anything'
 ]);

package/dist/db/backup.js

@@ -0,0 +1,28 @@
+import fsp from 'fs/promises';
+import path from 'path';
+import { SCAI_HOME } from '../constants.js';
+function getBackupDir() {
+    const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
+    return path.join(SCAI_HOME, `backup-${timestamp}`);
+}
+export async function backupScaiFolder() {
+    const backupDir = getBackupDir();
+    try {
+        await fsp.mkdir(backupDir, { recursive: true });
+        const files = await fsp.readdir(SCAI_HOME);
+        for (const file of files) {
+            const srcPath = path.join(SCAI_HOME, file);
+            const destPath = path.join(backupDir, file);
+            const stat = await fsp.stat(srcPath);
+            if (stat.isFile()) {
+                await fsp.copyFile(srcPath, destPath);
+            }
+        }
+        console.log(`📦 Backed up .scai folder to: ${backupDir}`);
+        return backupDir;
+    }
+    catch (err) {
+        console.warn('⚠️ Failed to back up .scai folder:', err instanceof Error ? err.message : err);
+        return null;
+    }
+}

package/dist/index.js

@@ -19,6 +19,7 @@ import { runFindCommand } from './commands/FindCmd.js';
 import { startDaemon } from './commands/DaemonCmd.js';
 import { runStopDaemonCommand } from "./commands/StopDaemonCmd.js";
 import { runAskCommand } from './commands/AskCmd.js';
+import { runBackupCommand } from './commands/BackupCmd.js';
 // 🎛️ CLI Setup
 const cmd = new Command('scai')
     .version(version)
@@ -96,14 +97,6 @@ cmd
     .action(() => {
     Config.show();
 });
-// Add explanation about alpha features directly in the help menu
-cmd.addHelpText('after', `
-🚨 Alpha Features:
-  - The "index", "daemon", "stop-daemon", "reset-db" commands are considered alpha features.
-  - These commands are in active development and may change in the future.
-
-💡 Use with caution and expect possible changes or instability.
-`);
 // 🔍 Indexing
 cmd
     .command('index [targetDir]')
@@ -112,6 +105,10 @@ cmd
     .action((targetDir, options) => {
     runIndexCommand(targetDir, { force: options.force });
 });
+cmd
+    .command('backup')
+    .description('Backup the current .scai folder')
+    .action(runBackupCommand);
 // 🧠 Query and assistant
 cmd
     .command('find <query>')
@@ -150,6 +147,14 @@ cmd
     }
     runModulePipelineFromCLI(file, options);
 });
+// Add explanation about alpha features directly in the help menu
+cmd.addHelpText('after', `
+🚨 Alpha Features:
+  - The "index", "daemon", "stop-daemon", "reset-db" commands are considered alpha features.
+  - These commands are in active development and may change in the future.
+
+💡 Use with caution and expect possible changes or instability.
+`);
 // ✅ Parse CLI args
 cmd.parse(process.argv);
 // 🔁 Apply global options post-parse

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "scai",
-  "version": "0.1.28",
+  "version": "0.1.30",
   "type": "module",
   "bin": {
     "scai": "./dist/index.js"