IndexArc Vault 2.0

Ultra-Secure, Portable AI Personal Vault & Intelligence Assistant

Features • Installation • AI Setup • Shortcuts • Security • Docs

---

Executive Overview

IndexArc is a zero-telemetry, portable personal vault and AI intelligence suite built for developers, engineers, and power users. It provides an all-in-one environment to store, organize, query, and refine sensitive API keys, tokens, environment configs, shell commands, and notes.

Unlike cloud password managers or traditional note apps, IndexArc runs 100% locally and stores all vault data next to the executable in a portable directory structure (data/ + config/). You can copy the entire folder onto a USB stick and run it on any computer without leaving traces in system directories or requiring administrator privileges.

---

Key Features & Capabilities

🔒 1. Secrets & Key Vault Management

• Multi-Extract Ingestion: Paste a blob of text (e.g. an entire .env file or terminal output) and IndexArc automatically parses out all embedded API keys, JWTs, AWS credentials, Telegram IDs, and tokens into individual structured candidates.
• Strict Naming & Freeform Types: Every secret requires a user-approved name and custom type (telegram user id, github PAT, AWS production secret).
• Unidentified Inbox: Candidates missing names or types are safely parked in the Needs Review inbox on Home until you label them.

✍️ 2. Advanced Scratchpad Notepad

• CSpell Trie Engine: Powered by cspell-lib and binary Arabic Tries (ar.trie.gz) with $O(L)$ Trie lookups.
• Developer Technical Dictionary: Includes developer terms (checkpointer, hardcoded, autocompletion, proofread, scratchpad, indexarc, etc.) to prevent false red underlines on code tokens.
• Custom Right-Click Context Menu:
  • Right-clicking any word under the cursor detects the token without needing manual selection.
  • Offers high-precision spelling suggestions.
  • Add "[word]" to Dictionary: Appends the word to config/user_dict.txt so it is permanently remembered across app restarts.
  • Ignore "[word]": Appends the word to config/ignored_words.txt so it is permanently ignored across app restarts.
• Inline Ghost-Text Predictions:
  • Copilot-style live text predictions appear as faint, italicized gray ghost text directly ahead of your typing cursor inside the text box.
  • Features an anti-duplication filter (extractContinuation) preventing text cloning across all LLM backends.
  • Press Tab ↹ or Right-Arrow to accept predictions, or Esc / type to dismiss.
• AI Rephrase & Proofread: Rewrite text across 6 tones (Human, Professional, Technical, Concise, Formal, Casual) or proofread with AI or automatic local CSpell Trie fallback.

🔍 3. Ask Intelligence & Hybrid Search

• Bilingual Natural Language Search: Query your vault in English or Arabic (e.g. "telegram id", "معرف تيليجرام", "bot token").
• Hybrid Search Engine: Combines BM25 keyword matching with Cosine similarity vector embeddings (VectorChunks).
• AI Summarized Answers: When an AI model is active, Ask generates a concise natural-language response directly referencing matching vault entries.

📁 4. Folder Scanner & Directory Watcher

• Multi-Format Ingestion: Recursively scans folders and extracts secrets from .txt, .env, .json, .docx (via Mammoth), .pdf (via PDF-Parse), .xlsx (via SheetJS), and source code files.
• Live Folder Watcher: Monitors designated directories using a debounced fs.watch pipeline and automatically flags newly added secrets in review sessions.

📚 5. Library & Duplication Management

• Filter Chips: Instant filtering by family (Secret, Command, Note, Unknown, Attention).
• Duplicate Finder: Identifies duplicate values or similar entry names across your vault and provides one-click bulk cleanup.

🔐 6. Bank-Grade Security & Portability

• AES-256-GCM Encryption: Optional master password protection using scrypt/Argon2id key derivation with 500ms timing-attack mitigation.
• Automatic Backups: Creates timestamped backups on every save (backups/vault-<stamp>.json), retaining the last 10 versions automatically.
• Zero-AppData Dependency: All configuration (config/) and data (data/) stay strictly inside the application root directory.

---

Installation & Setup Guide

Option A: Pre-compiled Desktop Releases (Windows)

Download the latest release from GitHub or use the pre-built executables:

Build Output	Path	Description
Portable Build	dist-desktop/IndexArc-Portable-2.0.0.exe	Zero installation. Double-click to run directly from USB or any folder.
Setup Installer	dist-desktop/IndexArc-Setup-2.0.0.exe	Standard Windows installer with Start Menu & Desktop shortcuts.

---

Option B: Running from Source (Developer Setup)

Prerequisites

• Node.js: Version 18.0.0 or higher
• npm: Version 9.0.0 or higher

Step 1: Clone Repository

git clone https://github.com/Mubder/IndexArc.git
cd IndexArc

Step 2: Install Dependencies

npm install

Step 3: Run Development Web Server

npm run dev

Open http://127.0.0.1:3000 in your browser. (Bound to 127.0.0.1 only).

Step 4: Run Electron Desktop App

npm run desktop

Step 5: Build Production Executables

# Build web production bundle & server
npm run build

# Package Windows portable & installer executables
npm run desktop:win

---

AI Engine Configuration

IndexArc supports 3 AI Operating Modes:

1. Local AI Mode (Ollama — 100% Offline & Private)

1. Install Ollama.
2. Download recommended models:

   ollama pull qwen2.5:0.5b
   ollama pull nomic-embed-text

3. Open IndexArc Settings tab and set AI Provider to Local (Ollama).

2. Cloud API Mode (Gemini, OpenAI, Groq, OpenRouter, Anthropic)

1. Get an API key from Google AI Studio, OpenAI, Groq, OpenRouter, or Anthropic.
2. Open IndexArc Settings tab, select your cloud provider, and paste your API key.
3. Keys are stored locally inside config/settings.json.

3. Heuristic Mode (No AI Required)

Without any AI provider configured, IndexArc's built-in regex & rule heuristic engine continues to parse .env files, API tokens, commands, and notes completely offline.

---

Folder Layout & Data Model

IndexArc/                     ← Copy this entire folder for USB portability
├── config/                   ← Application configuration
│   ├── settings.json         # AI provider, API keys, models, language
│   ├── user_dict.txt         # Persistent custom dictionary words
│   └── ignored_words.txt      # Persistent ignored words
├── data/                     # Vault data
│   ├── vault.json            # Encrypted or plaintext vault entries
│   ├── vectors.json          # Search embeddings
│   ├── watched_folders.json  # Monitored directory configurations
│   └── scan_sessions.json    # Folder scan session history
├── backups/                  # Automatic timestamped backups (last 10 retained)
├── dist/                     # Compiled web application and server bundle
├── docs/                     # Full technical documentation
├── electron-main.cjs         # Electron main process
├── server.ts                 # Express backend server
└── package.json

---

Keyboard Shortcuts & Controls

Shortcut	Context	Action
Tab ↹ or →	Scratchpad Editor	Accept inline ghost text prediction
Escape	Scratchpad Editor	Dismiss ghost prediction / Close context menu
Right-Click	Scratchpad Editor	Open Spelling Context Menu (Suggestions, Add, Ignore)
Ctrl + Z	Scratchpad Editor	Undo text edit
Ctrl + Y	Scratchpad Editor	Redo text edit

---

Environment Variables

Variable	Default	Purpose
INDEXARC_ROOT	Application directory	Override portable root directory location
GEMINI_API_KEY	None	API key override for development
PORT	3000	HTTP backend server port
HOST	127.0.0.1	Network binding interface (localhost only)

---

Documentation Index

Full technical documentation is located in the docs/ directory:

• 📖 User Guide — Comprehensive usage instructions, workflow guide, and AI setup.
• 🏗️ Architecture & Data Model — Process model, CSpell Trie engine, vector search, and encryption.
• 🔌 API Reference — Complete /api/* endpoint documentation.
• 💻 Developer Guide — Development workflow and extension guides.
• 📦 Build & Release Guide — Packaging, code signing, and distribution.
• 🛡️ Security Policy — Threat model, local encryption scheme, and Zero-Secrets compliance.

---

License & Security Policy

Private project. Built with strict Zero-Secrets Compliance — no private telemetry, tracking, or cloud uploads. All data remains exclusively under user control.