Skip to content

【third-party】Add Claude Code Skills for PaddleOCR Document Parsing and Text Recognition#17690

Merged
Bobholamovic merged 17 commits intoPaddlePaddle:mainfrom
Aidenwu0209:add-claude-code-skills
Mar 4, 2026
Merged

【third-party】Add Claude Code Skills for PaddleOCR Document Parsing and Text Recognition#17690
Bobholamovic merged 17 commits intoPaddlePaddle:mainfrom
Aidenwu0209:add-claude-code-skills

Conversation

@Aidenwu0209
Copy link
Contributor

Summary

  • Add two Claude Code skills for PaddleOCR API integration:
    • paddleocr-doc-parsing: Document parsing with layout detection, table extraction, and markdown output
    • paddleocr-text-recognition: OCR text recognition from images and PDFs
  • Each skill includes: API caller, configuration wizard, smoke test, optimization utilities
  • Shared .env configuration with PADDLEOCR_ACCESS_TOKEN

Changes

  • skills/paddleocr-doc-parsing/ — Document parsing skill (SKILL.md, scripts, references)
  • skills/paddleocr-text-recognition/ — Text recognition skill (SKILL.md, scripts, references)
  • .env.example — Environment variable template
  • README.md — Updated with skills section

Test plan

  • Run python scripts/configure.py to set up API credentials
  • Run python scripts/smoke_test.py to verify API connectivity
  • Test document parsing with sample PDF/image files
  • Test text recognition with sample images

🤖 Generated with Claude Code

@paddle-bot
Copy link

paddle-bot bot commented Feb 10, 2026

Thanks for your contribution!

…ition

Add two Claude Code skills integrating PaddleOCR APIs:

- paddleocr-doc-parsing: Document layout parsing with table extraction,
  formula recognition, and markdown output via PaddleOCR Document Parsing API
- paddleocr-text-recognition: OCR text recognition from images and PDFs
  via PaddleOCR Text Recognition API

Each skill includes: API caller, configuration wizard, smoke test,
file optimization utilities, and reference documentation.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@Aidenwu0209 Aidenwu0209 force-pushed the add-claude-code-skills branch from d0acca5 to 1ad68dc Compare February 10, 2026 08:56
Copy link
Member

@Bobholamovic Bobholamovic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

这几个skill是否可以考虑不针对Claude Code,而是更通用一些,其他的AI产品也能使用:

https://agentskills.io/specification

@Bobholamovic
Copy link
Member

另外可能需要检查一下本地开发环境是否安装了pre-commit钩子;代码风格检查好像没有通过

@Aidenwu0209 Aidenwu0209 force-pushed the add-claude-code-skills branch from ca44189 to 1ad68dc Compare February 12, 2026 05:16
Copy link
Member

@Bobholamovic Bobholamovic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

赞!感觉doc parsing的部分已经很接近最终版本了,留了一些意见辛苦看看。doc parsing的ready之后,可以参考doc parsing 更新一下text recognition部分

@Aidenwu0209 Aidenwu0209 force-pushed the add-claude-code-skills branch 3 times, most recently from 34d658c to e65ebee Compare February 13, 2026 03:49
@Aidenwu0209 Aidenwu0209 force-pushed the add-claude-code-skills branch from 39e1969 to f39934f Compare February 13, 2026 05:05
@Aidenwu0209 Aidenwu0209 force-pushed the add-claude-code-skills branch from 31f0237 to 1ba5b2f Compare February 14, 2026 03:30
Copy link
Member

@Bobholamovic Bobholamovic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

有一些新的建议

@Aidenwu0209
Copy link
Contributor Author

已修复 pre-commit 代码风格问题:

  1. CRLF → LF:所有文件已转换为 LF 行尾
  2. UTF-8 BOM:已移除多余的 BOM 标记
  3. black 格式化:已修复 4 个 Python 文件的格式问题
  4. flake8:检查通过(0 issues)

已在本地验证所有 pre-commit hook 检查通过。

@@ -0,0 +1,36 @@
# PaddleOCR Skills(中文)

本目录提供 PaddleOCR 云端 API 的 Agent Skills。
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

“PaddleOCR 云端 API” => “PaddleOCR 官方 API”

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done. Changed to「PaddleOCR 官方 API」.


## 仓库

- Source: <https://github.com/PaddlePaddle/PaddleOCR/tree/main/skills>
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

本身就在仓库里,应该就不用再专门这样写了

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done. Removed the 仓库/Repository section from both READMEs.

@@ -0,0 +1,36 @@
# PaddleOCR Skills(中文)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

不必标记“(中文)”

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done. Removed(中文)from the title.

Aidenwu0209 and others added 5 commits March 3, 2026 16:18
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Remove (中文) from Chinese README title
- Change 云端 API → 官方 API / cloud APIs → official APIs
- Remove 仓库/Repository section (already in repo)
- Remove UTF-8 BOM from both files

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add useDocUnwarping=False, useDocOrientationClassify=False, and
visualize=False to match the doc-parsing defaults.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Replace real-looking API key/token examples with obvious placeholders
- Recommend configure.py as primary setup method, .env as fallback

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Copy link
Member

@Bobholamovic Bobholamovic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@Bobholamovic Bobholamovic merged commit a310955 into PaddlePaddle:main Mar 4, 2026
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants