Foreword¶

When building educational Agents or RAG applications, a recurring issue is the lack of structured forms for original textbook content. Developers often have to parse PDFs or scrape web pages manually; versions and educational stages often mismatch, and scanned textbooks lack a text layer. Simply “finding the exact text from a specific chapter of a specific edition of 8th-grade physics” can consume a significant amount of time.

The Skill introduced below, “Middle School Physics Textbook Knowledge Base,” handles this task in advance: it distills all middle school physics PDF textbooks from the public GitHub repository TapXWorld/ChinaTextbook into an offline-searchable corpus. You can query chapter chunks and original excerpts by educational stage, grade, and version.

What is this¶

This is a community Skill on SkillHub, with coordinates @user_cefe6276/physics-textbooks, version 1.0.0, originating from the community, and published by user_cefe6276.

It solves a specific problem: allowing models or scripts to retrieve chapter chunks and original excerpts based on textbook dimensions—without opening PDFs or connecting to the internet—and obtaining download addresses for each textbook.

No API Key is required (tagged requires_api_key=false), and it falls under the knowledge-management category.

Core Features¶

1. Textbook Coverage

The data is derived from all middle school physics PDF textbooks in the public GitHub repository TapXWorld/ChinaTextbook, totaling 62 books: 22 for Junior High, 4 for the 5-4 school system, and 36 for Senior High.

2. Offline Retrieval

Retrieval is based on physics_corpus.json, which contains the table of contents, chapter chunks, and download addresses for each textbook. Search dimensions support locating chapter chunks and original excerpts by educational stage, grade, and version.

3. Download Links

Each textbook comes with domestic mirror and official repository download links, allowing for direct access when the original PDF is needed.

4. Source Traceability

The corpus is not compiled out of thin air; every chunk can be traced back to a specific textbook file within TapXWorld/ChinaTextbook.

Installation and Activation¶

The official page does not provide exclusive installation commands or configuration file paths, so this article will not make speculations.

Below is the general activation method for SKILL.md-based Skills on SkillHub; please refer to the official page for specifics:

Skill Coordinates: @user_cefe6276/physics-textbooks
Official Page: https://www.skillhub.cn/skills/user_cefe6276/physics-textbooks

Once activated, the Skill side provides retrieval capabilities based on physics_corpus.json. Note during use: this Skill is tagged requires_api_key=false, meaning no additional keys are needed, but also implying there are no configuration items for you to fill in.

Typical Usage¶

The official documentation does not provide reproducible invocation examples; the following only describes the corpus structure as a reference for integration.

Each record in physics_corpus.json contains three types of information:

Textbook Metadata (Stage / Grade / Version)
Chapter Table of Contents
Chapter Chunks + Original Excerpts + Download Address

Therefore, the typical retrieval logic is: first filter candidate textbooks by educational stage, grade, and version; then locate chunks within the candidate textbooks by chapter to extract original excerpts; if the original PDF is needed, use the domestic mirror or official repository links in the record to download.

During integration, it is recommended to run through the corpus first to confirm the completeness of chunks for the required stages and versions before connecting the upper-layer retrieval logic.

Applicable Scenarios and Notes¶

Suitable for:

Educational Agents / RAG applications that require original textbook text as a retrieval source
Offline tools that need to organize physics textbook corpora by educational stage, grade, and version
Scenarios requiring original textbook excerpts with traceable sources

Points to note:

The corpus scale covers 62 middle school physics textbooks and does not include other subjects.
The text from scanned textbooks comes from OCR or direct extraction; original excerpts may contain recognition errors. It is recommended to verify against the original PDF before direct citation.
The corpus originates from the public repository TapXWorld/ChinaTextbook; if upstream textbooks are updated, the chunks in this Skill need to be refreshed accordingly.
The data provided by the official page shows a rating of 100,000, 1,312 downloads, and 0 installations. The installation count of 0 indicates that there are currently no public integration cases; you need to verify the retrieval quality yourself before integration.

Conclusion¶

The value of this Skill lies in handling the manual labor of “finding textbooks, splitting chapters, and matching versions” in one go, providing an offline and traceable retrieval corpus. If you are building physics subject retrieval or Q&A capabilities, it is worth running it first to check the quality of the chunks.

Official Address: https://www.skillhub.cn/skills/user_cefe6276/physics-textbooks

Middle School Physics Textbook Corpus: Distilling 62 Middle School Physics Textbooks into an Offline Retrievable Corpus