intsig-textin/xparse-skills

34 stars Quality 44 Trend 34

Agent skills and CLI for document parsing powered by TextIn xParse API.

Overview

Agent skills and CLI for document parsing powered by TextIn xParse API. Turn PDFs, images, and Office documents into clean Markdown or structured JSON — directly inside your AI coding agent. Install into your agent with one command: Parse complete documents or navigate targeted sections, pages, facts, and tables via xparse-cli. PDF · Images (JPG/PNG/BMP/TIFF/WebP) · Word · PowerPoint · Excel · HTML · OFD · RTF - User provides a local file or document URL to read, convert, search, or extract content from - Task requires turning a document into agent-friendly text before further processing - Preparing content for summarization, analysis, or downstream workflows See SKILL.md for full routing rules, error handling, and references. xparse-cli is the underlying binary. It can also be used standalone.

README

xparse-skills

Agent skills and CLI for document parsing powered by TextIn xParse API.

Turn PDFs, images, and Office documents into clean Markdown or structured JSON — directly inside your AI coding agent.

Skills

Install into your agent with one command:

npx skills add intsig-textin/xparse-skills

xparse-parse

Parse complete documents or navigate targeted sections, pages, facts, and tables via xparse-cli.

Supported formats: PDF · Images (JPG/PNG/BMP/TIFF/WebP) · Word · PowerPoint · Excel · HTML · OFD · RTF

Use when:

User provides a local file or document URL to read, convert, search, or extract content from
Task requires turning a document into agent-friendly text before further processing
Preparing content for summarization, analysis, or downstream workflows

Quick start:

xparse-cli parse report.pdf --api auto                       # Markdown → stdout
xparse-cli parse report.pdf --api auto --view json           # Structured JSON output
xparse-cli parse report.pdf --api auto --output ./result/    # Save to directory

See SKILL.md for full routing rules, error handling, and references.

CLI

xparse-cli is the underlying binary. It can also be used standalone.

Install:

npm i -g xparse-cli

For users in China, use the npmmirror registry:

npm i -g xparse-cli --registry=https://registry.npmmirror.com

Key commands:

Command	Description
`xparse-cli parse`	Parse a document to Markdown or JSON
`xparse-cli quota --output json`	Show current daily free and authenticated free-package quota
`xparse-cli get_doc_info`	Create the stable local document ID for navigation
`xparse-cli get_outline`	Navigate the cached document outline
`xparse-cli search_text`	Search cached document content
`xparse-cli auth`	Configure API credentials (interactive)
`xparse-cli download --from result.json`	Download images from parse results
`xparse-cli version`	Show version info

Automatic, free, and paid API selection:

Mode	Behavior
`--api auto`	Default; uses current daily free quota, then an AppKey-authenticated `free_package.free_remain_count` when the service reports it
`--api free`	Force the free endpoint only
`--api paid`	Explicit paid route; requires user approval and valid OAuth/AppKey credentials

Run xparse-cli quota --output json for current authentication, allowance, page, and file-size limits instead of relying on hardcoded values.

Full CLI documentation: cli/README.md

Repository Structure

skills/
  xparse-parse/          # Agent skill definition and references
    SKILL.md
    references/
cli/                     # xparse-cli source (Go)
  cmd/
  internal/
  install/               # Install scripts
  build.sh               # Cross-compile script
.github/workflows/
  release.yml            # Auto-release on tag push

License

MIT

View this README on GitHub

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add intsig-textin/xparse-skills