发现优质的 AI Agent 技能

Extracting structured data from images, PDFs, and multi-turn form conversations.

This document describes how the baml-codegen skill queries the BoundaryML repositories via Model Context Protocol (MCP) servers to ensure fresh, validated patterns.

**Last Updated**: 2025-12-25 **Source**: BAML-REFERENCE-SOURCE.md (BoundaryML official gist), Hekmatica project, BoundaryML documentation

This document provides complete, real-world examples of BAML code generation for common use cases.

Complete parameter documentation for all 23 MCP tools.

学术论文数学公式转换工具 - 将Markdown中的LaTeX公式转换为docx和html格式

Clean controller architecture, validation strategy, error standardization, authentication & authorization, dan API documentation standards.

Log learnings and errors to markdown files for continuous improvement. Coding agents can later process these into fixes, and important learnings get promoted to project memory.

This document provides detailed examples for the ShapeScale-specific commands.

- Fast local parsing via PyMuPDF (`fitz`). - Less robust than specialized PDF parsers; table extraction is minimal. - `page.get_text("markdown")` gives quick Markdown output. - `page.get_text("text")` provides plain text for JSON. - Image extraction uses `page.get_images(full=True)` and `Pixmap`.

Parse PDFs locally using PyMuPDF for fast, lightweight extraction into Markdown by default, with optional JSON and image/table outputs in a per-document directory.

Parse a PDF locally with MinerU (CPU). Default output is Markdown + JSON. Use tables/images only when requested.