Download the PHP package blockshiftnetwork/chat-markdown-converter without Composer

Chat Markdown Converter

Convert AI-generated Markdown to WhatsApp, Telegram, Discord, Slack and Instagram formats using an Intermediate Representation (IR). Perfect for converting LLM responses to chat-friendly formats.

This PHP library transforms ChatGPT, Claude, GPT-5, and other AI model outputs into platform-specific markup. It handles code blocks, tables, lists, links, and rich text formatting while maintaining readability across all supported platforms.

Features

• Fluent API with chainable method calls
• Clean architecture using the Intermediate Representation pattern for extensibility
• Comprehensive test coverage with 200+ passing tests (Pest)
• Platform-specific rendering optimized for each chat platform
• Smart message chunking that splits text at safe breakpoints
• Zero external dependencies, lightweight implementation

Use Cases

AI Chatbots & Virtual Assistants

Send formatted responses from OpenAI, Anthropic, or other LLM APIs directly to users via their preferred messaging platform. Ensure code blocks, tables, and lists render correctly across Telegram, WhatsApp, Discord, Slack, and Instagram.

Customer Support Automation

Automate support workflows by converting AI-generated help articles and documentation into chat-friendly formats. Preserve formatting while delivering concise, readable responses in your customers' channels.

Developer Tools & DevOps

Integrate with CI/CD pipelines, monitoring systems, or alerting platforms to send formatted logs, error messages, or status updates to team channels. Convert Markdown reports to platform-appropriate syntax automatically.

Content Distribution Systems

Distribute newsletters, summaries, or generated content across multiple platforms simultaneously. Write once in Markdown and automatically convert to Telegram HTML, WhatsApp text, Discord markdown, Slack mrkdwn, or Instagram-ready Unicode captions.

Educational Platforms

Convert AI-generated tutorials, code examples, and learning materials into appropriate formats for students across different communication channels. Keep code snippets and syntax highlighting functional.

Knowledge Base Integration

Transform documentation and wiki content from Markdown sources into chat-appropriate formats for quick distribution via messaging apps. Maintain structure and readability without manual reformatting.

Notification Systems

Build notification services that send formatted alerts, summaries, and reports from AI analysis tools. Convert complex data presentations into simple, readable messages on any platform.

Laravel Applications

Use the package inside Laravel jobs, notifications, controllers, or queued bot handlers when an application needs to deliver AI-generated Markdown through chat channels without manually rewriting the same response for each platform.

Multi-Platform Content Management

Manage content in a single Markdown format while automatically adapting output for different chat platforms' unique formatting requirements and limitations.

Internal Communication Tools

Streamline internal updates, announcements, and reports by converting Markdown-based company communications to team messaging platforms with proper formatting.

Installation

You can install the package via composer:

Why Choose This Library?

Platform-Aware Formatting

Unlike naive Markdown-to-text converters, this library understands each platform's unique limitations and formatting rules. Telegram uses HTML tags, WhatsApp uses asterisk-based formatting, Discord and Slack have their own markdown variants, and Instagram requires Unicode Mathematical Alphanumeric Symbols since it has no native markup at all - we handle all these differences automatically.

Intermediate Representation Architecture

By parsing Markdown into an abstract IR first, we ensure consistent behavior across all platforms. This clean architecture makes it easy to add new platforms or customize rendering logic without modifying the core parser.

Intelligent Feature Adaptation

Automatically converts complex Markdown features to platform-compatible formats:

• Tables become bullet lists on platforms without native table support
• Code blocks are preserved with proper language tags
• Task lists adapt to each platform's checkbox syntax or emoji alternatives
• Links transform to each platform's expected format

Production-Ready Reliability

Comprehensive test coverage (200+ tests) ensures consistent behavior across edge cases. Handle special characters, nested formatting, mixed content types, and Unicode/emoji support with confidence.

Developer Experience

Simple, intuitive API with fluent method chaining. Convert in one line with static methods or take full control with the flexible parser options. Zero learning curve for Markdown developers.

Performance Optimized

Lightweight implementation with no external dependencies. Fast parsing and rendering suitable for high-volume applications and real-time chat systems.

AI-Response Ready

Specifically designed for LLM outputs that often include code blocks, tables, lists, and mixed formatting. Preserve the intelligence of AI responses while making them platform-appropriate.

Open Source & Extensible

MIT license, active development, and contribution-friendly architecture. Easily create custom renderers for proprietary platforms or specialized use cases.

Alternative Approaches

Direct Markdown-to-Text Conversion

Simple converters often strip all formatting or produce unreadable output. Our platform-specific renderers preserve structure and readability while respecting each platform's formatting capabilities.

Multiple Separate Libraries

Using different libraries for each platform adds complexity and dependencies. Our unified solution provides consistent behavior across all platforms with a single, well-tested codebase.

Manual Reformatting

Manually writing separate responses for each platform is time-consuming and error-prone. Automate with our converter and maintain content consistency while reaching users everywhere.

Regular Expression Replacements

Naive regex-based converters fail with nested formatting, special characters, and complex structures. Our parser handles all edge cases correctly with proper state management.

Usage

Quick Conversion

In a Laravel Notification or Job

Fluent API

With Message Chunking

Custom Renderer

Parser Options

Supported Features

Currently Supported

• Text Formatting: Bold, italic, strikethrough, inline code, highlight
• Headers: Markdown headings (# ## ###)
• Code Blocks: Code blocks with language preservation
• Task Lists: Checkbox support with emoji conversion (WhatsApp)
• Links: Markdown link formatting
• Images: Image formatting
• Lists: Numbered and bullet lists
• Blockquotes: Quote blocks
• Tables: Auto-converted to bullet lists for non-table platforms
• Message Chunking: Smart text splitting with safe breakpoints
• Unicode Support: Full UTF-8 support including emojis
• Multiple Platforms: Telegram, WhatsApp, Discord, Slack, Instagram

Platform-Specific Features

Telegram (HTML Mode)

• Bold: <b>text</b>
• Italic: <i>text</i>
• Strikethrough: <s>text</s>
• Highlight: <b>text</b>
• Inline Code: <code>text</code>
• Code Blocks: <pre>code</pre>
• Headers: <b>text</b> (bold)
• Links: <a href="url">text</a>
• Images: <a href="url">text</a> (without !)
• Task Lists: - [x] and - [ ] (native support)

WhatsApp

• Bold: *text* (single asterisk)
• Italic: _text_ (underscore)
• Strikethrough: ~text~ (single tilde)
• Highlight: *text* (bold)
• Inline Code: `text` (backticks)
• Code Blocks: Triple backticks
• Headers: *text* (bold, replaces # syntax)
• Links: text: url (colon separator)
• Images: text: url (without !)
• Task Lists: ✅ task (completed) and ⬜ task (pending) with emojis

Discord

• Bold: **text**
• Italic: *text* (asterisks only, not underscores)
• Strikethrough: ~~text~~
• Highlight: **text** (bold)
• Inline Code: `text` (backticks)
• Code Blocks: Triple backticks with language
• Headers: **text** (bold, replaces # syntax)
• Links: [text](url) (markdown format)
• Images: [text](url) (without !)
• Task Lists: - [x] and - [ ] (native support)

Slack

• Bold: *text* (single asterisk)
• Strikethrough: ~text~ (single tilde)
• Highlight: *text* (bold)
• Inline Code: `text` (backticks)
• Code Blocks: Triple backticks with language
• Headers: *text* (bold, replaces # syntax)
• Links: <url|text> (Slack format)
• Images: <url|text> (without !)
• Task Lists: - [x] and - [ ] (native support)

Instagram

Instagram has no native rich-text formatting in any surface (captions, bios, comments, DMs, story overlays, Reels descriptions). The renderer substitutes ASCII letters and digits with characters from Unicode's Mathematical Alphanumeric Symbols block so the output renders with the expected visual weight when pasted into Instagram.

• Bold: Sans-serif bold Unicode glyphs (e.g. b becomes 𝗯)
• Italic: Sans-serif italic Unicode glyphs (digits remain plain — Unicode has no italic digits)
• Bold Italic: Sans-serif bold-italic Unicode glyphs
• Strikethrough: Combining long stroke overlay U+0336 after each character
• Highlight: Sans-serif bold (Instagram has no highlight equivalent)
• Inline Code: Monospace Unicode glyphs (e.g. c becomes 𝚌)
• Code Blocks: Monospace body wrapped above and below with a heavy horizontal rule (━━━━━━━━━━━━━━━━); triple-backticks paste verbatim on Instagram so they're not used
• Headers: Sans-serif bold (regardless of # level)
• Links: text: url (Instagram captions don't auto-link http(s):// URLs)
• Images: alt: url (the ! prefix is stripped)
• Blockquotes: ❝ quote ❞ (typographic quotation marks)
• Horizontal Rule: ━━━━━━━━━━━━━━━━ (heavy box-drawing line)
• Task Lists: ✅ (completed) and ⬜ (pending) with emojis
• Bullet Lists: Leading - replaced with • for visual polish
• Tables: Converted to bullet points with bold headers (Instagram has no table support)

Accessibility note: Mathematical Alphanumeric Symbols are read as separate, decontextualized characters by screen readers (e.g. 𝗯𝗼𝗹𝗱 is announced as "Mathematical Sans-Serif Bold B, Mathematical Sans-Serif Bold O…"). When accessibility matters more than visual emphasis, post the original Markdown body without running it through the Instagram renderer.

Roadmap

Completed

• Platform-specific link formatting
• Table to bullet conversion
• Headers support
• Highlight syntax
• Image formatting improvements
• Nested/complex formatting

Planned

• Telegram MarkdownV2 support
• Additional platforms (Teams, Mattermost, Signal)

Platform Comparison

Testing

Current Test Status: 206 passed, 1 skipped

Architecture

The package uses an Intermediate Representation (IR) pattern:

Components

• Parser: Converts Markdown to IR
• HeaderParser: Detects and parses markdown headers (# ## ###)
• Parsers: Specialized parsers for code blocks, tables, links, styles, blockquotes, horizontal rules
• Renderers: Platform-specific renderers (Telegram, WhatsApp, Discord, Slack, Instagram)
• Support: IR, TextChunker

Contributing

Contributions are welcome! Please see CONTRIBUTING for details.

Development Setup

Credits

• Blockshift Network
• All Contributors

FAQ

Why use an Intermediate Representation instead of direct conversion?

The IR pattern provides better separation of concerns and extensibility. Parse once, render many times. This makes it easy to add new platforms without modifying the parser logic, and ensures consistent behavior across all renderers.

Can I customize the rendering behavior?

Yes! Extend the AbstractRenderer class to create custom renderers. The parser provides a structured IR that you can transform into any format you need.

How does table conversion work?

For platforms without native table support (Telegram, WhatsApp, Slack, Instagram), tables are automatically converted to hierarchical bullet lists, preserving the structure and readability.

Does this library support all Markdown features?

We support the most common Markdown features used in AI responses: headings, code blocks, lists, links, images, blockquotes, horizontal rules, and text formatting. See the Supported Features section for details.

What about message length limits?

The TextChunker intelligently splits long messages at safe breakpoints (after sentences, paragraphs, or list items) while preserving formatting and avoiding broken markup. To land on a clean breakpoint, a chunk may exceed the requested maxLength by up to 50 characters; budget for that overshoot when choosing your limit (e.g. pass maxLength: 4046 if you absolutely must stay under Telegram's 4096-char hard cap).

Can I use this with any PHP version?

This library requires PHP 8.3 or higher, taking advantage of modern PHP features like match expressions and readonly properties.

Is this suitable for production use?

Yes! The library has comprehensive test coverage (200+ tests) and is actively maintained. It's designed for performance and reliability in production environments.

Optimization Tips

• Reuse parsers: Create a single parser instance and reuse it for multiple conversions
• Batch operations: Process multiple messages in a single request when possible
• Cache rendered output: Store converted results for frequently used content
• Disable unnecessary features: Use parser options to skip parsing for content that doesn't need certain features

License

The MIT License (MIT). Please see License File for more information.