Welcome to the Data Rescue Recovery Intelligence Suite — a pioneering software platform designed to salvage, reconstruct, and protect digital assets from accidental deletion, corruption, or system failures. Unlike conventional recovery tools, Data Rescue employs a multi-layered heuristic engine that mimics forensic data reconstruction, offering a lifeline for mission-critical files. This repository contains the full source code, documentation, and community-driven enhancements for version 5.2.0 (2026 release).
Important: This project focuses on legitimate data salvage methodologies. All described techniques adhere to ethical computing standards and legal data recovery practices. No unauthorized access methods are included.
- Quick Start — Download & Install
- System Requirements & OS Compatibility
- Architecture Overview
- Key Features & Capabilities
- Example Profile Configuration
- Example Console Invocation
- API Integrations: OpenAI & Claude
- Responsive UI & Multilingual Support
- 24/7 Customer Support & Community
- Mermaid Diagram: Recovery Pipeline
- License Information
- Disclaimer & Ethical Use
Step 1: Click the download badge above to retrieve the latest release package (v2026.2.1).
Step 2: Extract the archive to a dedicated directory (e.g., C:\DataRescue\ or ~/data-rescue/).
Step 3: Run the installer script (install.sh for Linux/macOS, setup.exe for Windows).
Step 4: Follow the on-screen prompts to complete the configuration wizard.
💡 Pro Tip: For enterprise deployments, use the silent install flag:
--silent --license-key [PRODUCT_KEY]to streamline batch deployments across multiple workstations.
| Operating System | Version | Status (2026) | Emoji |
|---|---|---|---|
| Windows 11 | 23H2+ | ✅ Fully Supported | 🪟 |
| Windows 10 | 22H2+ | ✅ Fully Supported | 🪟 |
| macOS Sonoma | 14.x+ | ✅ Fully Supported | 🍎 |
| macOS Sequoia | 15.x+ | ✅ Supported (Beta) | 🍎 |
| Ubuntu | 24.04 LTS | ✅ Fully Supported | 🐧 |
| Debian | 12+ | ✅ Fully Supported | 🐧 |
| Fedora | 40+ | ✅ Supported | 🐧 |
| Android (Termux) | 14+ | 🤖 |
Minimum Requirements:
- CPU: 2.0 GHz dual-core (x86_64 / ARM64)
- RAM: 4 GB (8 GB recommended for deep scans)
- Disk: 500 MB free space (plus recovery target storage)
- Display: 1280×720 resolution
Data Rescue operates as a modular pipeline with four core layers:
- Scan Engine — Low-level block-by-block analysis using direct disk I/O (Windows
\\.\PHYSICALDRIVE*, macOS/dev/rdisk*, Linux/dev/sd*). - Signature Database — Over 2,500 known file signatures (magic bytes) for identifying image, video, document, and database formats.
- Reconstruction Module — De-fragments and reassembles file chunks using temporal and spatial heuristics.
- Export Wrapper — Outputs recovered files to user-specified directories with optional hashing and metadata preservation.
Each layer communicates via a lightweight IPC bus (ZeroMQ v4.3+), enabling distributed recovery across networked nodes.
- Deep Sector Scanning — Recovers files from formatted, partially overwritten, or partitioned drives.
- Forensic Mode — Generates chain-of-custody logs with SHA-256 hashes for legal admissibility.
- Selective Restoration — Preview files before recovery; recover only what you need.
- RAID Reconstructor — Supports JBOD, RAID 0/1/5/6/10 with automatic parity calculation.
- Responsive UI — Adaptive interface scales from 4K monitors to 7-inch tablets.
- Multilingual Support — Fully localized in 18 languages including English, Spanish, Mandarin, Arabic, and Hindi.
- 24/7 Customer Support — Real-time chat, email ticketing, and scheduled remote assistance.
- Automated Backups — Schedule weekly integrity scans with email alerts.
- OpenAI API Integration — Uses GPT-4o to analyze file structures and suggest recovery strategies.
- Claude API Integration — Leverages Claude 3.5 Sonnet for natural-language reporting of scan results.
- Zero Network Dependency — All core functions operate offline; API features are optional.
- Portable Mode — Run directly from a USB stick without installation.
The profiles/default.yaml file defines recovery parameters. Below is a sample configuration optimized for document recovery from an SSD:
profile_name: "ssd_document_highspeed"
target_device: "/dev/sdb"
scan_mode: "quick" # options: quick, deep, forensic
file_types:
- pdf
- docx
- xlsx
- pptx
- odt
output_directory: "/mnt/recovery_2026/"
preserve_timestamps: true
generate_manifest: true
compact_lost_chunks: false
max_concurrent_threads: 16
signature_overrides:
- extension: "custom_doc"
magic_bytes: "0x89504E47"
mimetype: "application/x-custom-doc"
api_integration:
openai_model: "gpt-4o"
claude_model: "claude-3-5-sonnet-20241022"
auto_suggest_filters: trueExplanation: This profile uses the quick scan mode to target recent documents on an SSD, limits thread count to avoid overheating, and enables both AI APIs for intelligent reconstructions.
Basic Recovery (Linux/macOS):
./data-rescue recover --device /dev/sdc --profile profiles/default.yamlAdvanced Recovery with AI Help (Windows PowerShell):
.\data-rescue.exe recover `
--device \\.\PHYSICALDRIVE2 `
--profile profiles/default.yaml `
--output D:\recovered `
--enable-openai `
--enable-claude `
--verboseDry Run (Preview Only):
./data-rescue recover --device /dev/sdd --preview-only --output /tmp/previewOutput Example:
[INFO] Scan engine initialized (v2026.2.1)
[INFO] Device: /dev/sdc (931.51 GB, sector 512B)
[INFO] Profile loaded: ssd_document_highspeed
[INFO] Detected 1,247 recoverable files (245 MB)
[INFO] AI analysis suggested 13 file merges (signature mismatch)
[INFO] Recovery started at 14:32:01 UTC
[PROGRESS] [====================] 100%
[INFO] Recovery completed in 47 seconds
[INFO] 1,233 files written to /mnt/recovery_2026/
[INFO] Manifest saved to recovery_2026_manifest.json
Data Rescue features optional integration with two leading AI platforms:
- Purpose: Analyze fragmented data to infer original file structure.
- Example Use: When the scan engine cannot identify a file signature, GPT-4o reviews surrounding binary patterns and suggests a plausible format (e.g., "This appears to be a corrupted JPEG with missing EXIF header — probabilities: 87% JPEG, 9% TIFF, 4% unknown").
- Setup: Set
OPENAI_API_KEYin environment variables.
- Purpose: Generate human-readable recovery reports and documentation.
- Example Use: After scanning, Claude produces a plain-English summary: "We recovered 45 spreadsheets from a partially overwritten NTFS volume. 3 files were truncated and might require manual repair in Excel."
- Setup: Set
ANTHROPIC_API_KEYin environment variables.
Note: Both integrations are entirely optional. The core recovery engine operates without any network connectivity.
Responsive UI
The graphical interface is built using Qt 6.7 with a vector-based layout that adapts to screen resolutions from 1024×768 to 3840×2160. On mobile devices (tablets running Android or iPadOS via companion app), the interface collapses into a single-pane workflow with touch-optimized controls.
Multilingual Support (18 Languages)
- English (default)
- 中文 (简体)
- Español
- العربية
- हिन्दी
- Français
- Português
- Русский
- 日本語
- Deutsch
- Italiano
- 한국어
- Türkçe
- Tiếng Việt
- Bahasa Indonesia
- Nederlands
- Polski
- Svenska
Language detection occurs automatically based on system locale, or can be overridden via --lang flag.
Support Channels (all free of charge):
- Discord Server: Real-time help from developers and power users.
- GitHub Issues: Feature requests, bug reports, and diagnostics.
- Email Ticketing: Response within 2 hours (business days) or 8 hours (weekends/holidays).
- Remote Assistance: Scheduled TeamViewer/AnyDesk sessions for complex recoveries.
Community Contributions
We encourage pull requests for:
- New file signature definitions (submit
signatures/new_format.json) - Translation updates (see
lang/directory) - Recovery algorithm optimizations (submitted as PRs with benchmarks)
graph TD
A[Damaged Storage Device] --> B[Block-Level Scanner]
B --> C{Signature Match?}
C -->|Yes| D[Extract Known Format]
C -->|No| E[Heuristic Analysis]
E --> F[AI-Assisted Pattern Recognition]
F --> G[OpenAI / Claude Inference]
G --> H{Contiguous Data?}
H -->|Yes| I[Reassemble File]
H -->|No| J[Fragment Manager]
J --> K[De-fragmentation Engine]
K --> I
I --> L[Integrity Check]
L --> M{Passes Check?}
M -->|Yes| N[Export to Output Directory]
M -->|No| O[Mark as Corrupted + Log]
N --> P[Manifest Generation]
O --> P
P --> Q[User Notification UI]
Q --> R[Recovery Complete]
This project is licensed under the MIT License — a permissive open-source license that allows free use, modification, and distribution of the software, provided the original copyright notice and disclaimer are preserved.
Summary:
- ✅ Commercial use allowed
- ✅ Modification allowed
- ✅ Distribution allowed
⚠️ Liability limitation (no warranty)⚠️ Must include original copyright notice
Copyright © 2026 Data Rescue Contributors
Data Rescue is a tool for legitimate data recovery from devices you own or have explicit written permission to access. The software does not facilitate unauthorized access to third-party systems, nor does it include any backdoors, keyloggers, or spyware.
By using this software, you agree to:
- Recover data only from devices for which you hold lawful ownership or explicit authorization.
- Comply with all applicable local, national, and international laws regarding data privacy and electronic communications.
- Assume full responsibility for any data loss, corruption, or legal consequences arising from misuse.
The developers and contributors are not liable for:
- Data recovered from devices without lawful authorization.
- Use of the software in violation of the Digital Millennium Copyright Act (DMCA) or equivalent legislation.
- Any indirect, incidental, or consequential damages arising from software usage.
Ethical Recovery Pledge:
We do not condone the use of this software for retrieving stolen, classified, or unlawfully obtained data. All recovery operations must be conducted transparently and with proper consent.
Thank you for choosing Data Rescue — because your digital memories deserve a second chance.