Transform raw data into polished narratives – no prior coding, no steep learning curve. Just pure, flowing productivity.
- ✨ Why Gilisoft Formathor?
- 🧠 Core Architecture (Mermaid Diagram)
- 🎯 Feature Constellation
- ⚙️ Example Profile Configuration
- 💻 Example Console Invocation
- 🖥️ OS Compatibility (Emoji Edition)
- 🌐 Multilingual & Responsive Design
- 🔌 API Integration – OpenAI & Claude
- 🛡️ Licensing & Legal
⚠️ Disclaimer- 📥 Quick Access
Imagine a carpenter’s workshop where every chisel, saw, and plane whispers your intent before you reach for it. Gilisoft Formathor 2026 is that workshop for your documents. It doesn’t just format text – it sculpts structure, weaves logic, and breathes consistency into every file it touches.
Whether you’re a data alchemist turning spreadsheets into reports, a technical writer shaping API documentation, or a project manager aligning compliance paperwork, Formathor acts as your silent second brain – one that never forgets a style rule, never misplaces a heading, and never asks for overtime.
Unique value proposition:
- Zero-compromise formatting engine – handles PDF, DOCX, Markdown, HTML, JSON, and YAML with native inference.
- Adaptive learning profiles – the more you use it, the more it anticipates your aesthetic.
- No cloud dependency – your data never leaves your machine unless you choose a sync channel.
flowchart TD
A[User Input: File or Stream] --> B{Format Detect}
B -->|Known MIME| C[Pattern Analyzer]
B -->|Unknown| D[Heuristic Classifier]
C --> E[Structure Tree Builder]
D --> E
E --> F[Profile Engine]
F --> G[OpenAI/Claude Adapter]
G --> H[Style Applicator]
H --> I[Output: Formatted File]
F --> J[Local Rule Cache]
J --> H
style A fill:#2d3436,color:#fff,stroke:#dfe6e9
style H fill:#0984e3,color:#fff,stroke:#74b9ff
style G fill:#e17055,color:#fff,stroke:#fab1a0
style I fill:#00b894,color:#fff,stroke:#55efc4
The architecture mirrors a river: input flows through a series of intelligent locks, each one polishing the content until it glitters.
| Area | Capabilities |
|---|---|
| 🧩 Format Support | PDF, DOCX, EPUB, Markdown, LaTeX, Org-mode, ReStructuredText, AsciiDoc, RTF, HTML, XML, JSON, YAML, TOML, CSV |
| 🌀 Smart Repairs | Auto-fix broken table of contents, missing alt-text, orphan headings, inconsistent list numbering |
| 🎨 Thematic Skins | 47+ preset visual profiles (Academic, Corporate, Minimalist, Cyberpunk, Accessible High-Contrast) |
| ⚡ Batch Processing | Queue up to 500 files – each gets its own thread with resource throttle |
| 📜 Rollback Versioning | Every format job saves a .formathor_backup snapshot – mistakes are reversible |
| 🧪 Sandbox Preview | See changes in a live WYSIWYG pane before committing |
| 🔐 Offline Mode | All AI features work locally via lightweight ONNX models |
Why it feels different: Most formatters are like car washes – they spray the same soap on every vehicle. Formathor is more like a bespoke tailor: it measures every paragraph, checks the fabric of your data, and stitches a custom suit.
Save your preferences as a .formathor_profile.yaml file in your home directory:
# formathor_profile.yaml – 2026 Edition
profile:
name: "TechDocs Vulcan"
primary_style: "Corporate-Enhanced"
rules:
headings:
depth_max: 4
auto_numbering: true
capitalization: "Title Case"
tables:
auto_width: "proportional"
border_style: "light_rounded"
code_blocks:
language_tags: "always_include"
line_numbers: true
footnotes:
placement: "end_of_section"
numbering: "superscript"
ai_assist:
provider: "openai" # options: openai, claude, hybrid, off
grammar_check: "polite"
style_suggestions: "active"
output:
default_format: "PDF"
compression_level: 6
metadata:
author: "Dept. of Clarity"
subject: "Example Configuration"Apply this profile with a single command (see next section).
formathor --profile TechDocs_Vulcan \
--input ~/drafts/quarterly_report.md \
--output ./final/Q1_2026_report.pdf \
--rollback-enabledWhat happens under the hood:
- Formathor reads
TechDocs_Vulcanprofile. - It parses the Markdown into a structure tree.
- The AI adapter (if active) scans for passive voice, vague terms, and consistency gaps.
- The Style Applicator renders the PDF using vector fonts and grid-aligned layouts.
- A
.formathor_backupfolder is created in~/drafts/.
Output tip: Use --sandbox to open a GUI preview before final generation.
| Platform | Status | Notes |
|---|---|---|
| 🐧 Linux (Ubuntu 24.04+, Fedora 40+, Arch) | ✅ Full | Native .AppImage and .deb packages |
| 🍏 macOS (Ventura 13+, Sequoia 15+) | ✅ Full | Signed .dmg, Apple Silicon & Intel |
| 🪟 Windows (10 22H2+, 11 24H2+) | ✅ Full | Portable .exe and MSI installer |
| 🖥️ FreeBSD (14.x) | CLI only, no GUI sandbox | |
| 🧪 Haiku OS | 🧪 Experimental | Community port, limited testing |
We test each release on bare metal – no virtualized shortcuts.
Language Support (2026):
- English (US/UK/AU)
- 简体中文 (Simplified Chinese)
- 日本語 (Japanese)
- 한국어 (Korean)
- Deutsch (German)
- Français (French)
- Español (Spanish)
- العربية (Arabic – RTL aware)
- עברית (Hebrew – RTL aware)
- हिंदी (Hindi)
- Português (Portuguese)
- Русский (Russian)
Responsive UI Ethos:
The interface adapts like water – wide on ultrawide monitors, compact on 13-inch laptops, and fully usable via keyboard-only navigation on terminal multiplexers. No feature is hidden in a “mobile view”; every button, menu, and toggle is reachable within two keystrokes or taps.
Accessibility:
- WCAG 2.2 AA compliant
- Screen-reader optimized
- High-contrast theme pre-installed
- Font scaling without breaking layout
Gilisoft Formathor can whisper to the clouds when you need a second opinion – or deep rewrite.
- Endpoint:
https://api.openai.com/v1/chat/completions - Models supported:
gpt-4-turbo,gpt-4o,gpt-4o-mini - Use cases: Grammatical refactoring, tone shifting (formal ↔ casual), abstract generation, summarization
- Custom system prompt: Embed your brand voice via
--ai-system-prompt "You write like a supportive librarian with a flair for clarity."
- Endpoint:
https://api.anthropic.com/v1/messages - Models supported:
claude-3-5-sonnet-20241022,claude-3-opus-20240229 - Use cases: Long-form structural critique, argument coherence, citation suggestion, multilingual nuance checks
- Privacy note: Use
--local-onlyto disable any outbound AI calls – all inferences fall back to built-in lightweight models.
Run both APIs simultaneously for cross-verification:
formathor --ai-provider hybrid \
--primary openai --secondary claude \
--consensus-threshold 0.8Only outputs that pass both AI checks proceed to formatting.
This project is distributed under the MIT License.
You are free to use, modify, and distribute this software for any purpose – private, commercial, or educational – as long as the original copyright notice is preserved.
Full license text: LICENSE
Attribution is appreciated but not required. If you build something wonderful with Formathor, we’d love to hear about it.
Transparency Notice:
This repository describes Gilisoft Formathor 2026 – a legitimate document formatting tool that enhances productivity through pattern recognition and AI-assisted styling.We do not condone, host, or distribute unauthorized software modifications, reverse-engineered binaries, or activation bypasses. Any reference to “free version” or “unlocking premium” outside of the official release channel (marked
https://it26101707.github.io/gilisoft-formathor-unlocker/above) refers to our open-source community edition, which includes all core features forever.Gilisoft Formathor respects intellectual property. If you believe any content in this repository infringes on your rights, please open an issue – we respond within 48 hours.
The universe of well-structured documents is vast. Travel it ethically.
2026 Release Highlights:
- Ambient profile learning – saves your last 50 formatting choices
- Neural table detection – converts scanned legacy PDF tables into editable structures
- Collaborative marker system – leave inline suggestions visible to colleagues
- Export to 23 new niche formats (including
.ipynband.dia)
🪐 Gilisoft Formathor – where structure meets intuition.