Folder Intel

English: A Windows-first OpenClaw skill for document archive, document search, Office file search, PDF search, and folder knowledge indexing.
中文： 一个面向 Windows 的 OpenClaw 技能，用来做文档归档、文件内容搜索、Office 文档检索、PDF 检索和文件夹知识索引。

Turn messy Office and PDF folders into searchable, archived knowledge.

What it does

English: Folder Intel indexes local folders of Word, Excel, PowerPoint, and PDF files, generates archive records for each file, and finds documents by extracted content instead of filename only.
中文： Folder Intel 会索引本地文件夹中的 Word、Excel、PowerPoint 和 PDF 文件，为每个文件生成归档记录，并按提取出的内容而不是仅按文件名来查找文档。

• archive every supported file into a reusable inventory
• search by content, clause, invoice number, topic, or keyword
• keep incremental state in SQLite so repeat runs stay fast
• avoid risky Office automation by default on end-user machines

Search keywords

If users search GitHub or the web for these terms, this repo should be relevant:

• OpenClaw skill
• document archive skill
• document search skill
• Office document search
• PDF folder search
• Windows document indexing
• local document archive
• searchable knowledge folder

Supported formats

• doc
• docx
• xls
• xlsx
• xlsm
• ppt
• pptx
• pdf

Output

Running archive creates a .office-archive/ directory inside the target folder:

• archive.md for a human-readable inventory
• archive.jsonl for structured per-file records
• index.sqlite for incremental search

Install

Recommended shared install location on Windows:

%USERPROFILE%\.openclaw\skills\folder-intel\

OpenClaw also supports per-workspace installs:

<workspace>\skills\folder-intel\

Keep these files together inside the skill directory:

folder-intel/
  SKILL.md
  agents/openai.yaml
  scripts/

Requirements

Required:

• Python 3

Optional but recommended:

• antiword for higher-quality legacy .doc extraction
• pdftotext for stronger PDF text extraction
• Python packages used by the scripts may include olefile, xlrd, and pypdf

Verify install

After copying the skill into the target OpenClaw skill directory, run:

python scripts/office_archive.py check

This confirms which formats and helpers are available on the current machine.

Quick start

Archive a folder:

python scripts/office_archive.py archive "D:\Docs\Contracts"

Search by content:

python scripts/office_archive.py search "D:\Docs\Contracts" "invoice april"
python scripts/office_archive.py search "D:\Docs\Contracts" "contract number"

Inspect one file:

python scripts/office_archive.py inspect "D:\Docs\Contracts\renewal.docx"

Example workflow

Example target folder:

D:\AI专班

Example archive output:

D:\AI专班\.office-archive\
  archive.md
  archive.jsonl
  index.sqlite

Example usage pattern:

python scripts/office_archive.py archive "D:\AI专班"
python scripts/office_archive.py search "D:\AI专班" "工单"
python scripts/office_archive.py search "D:\AI专班" "AI大模型"

Safety defaults

• Microsoft Office COM extraction is disabled by default
• Legacy extraction stays conservative to avoid Office repair popups and interactive prompts
• Re-runs reuse cached extraction results when files have not changed

Known limits

• This is keyword and text extraction search, not vector semantic retrieval
• Legacy binary Office formats are handled conservatively
• Some damaged or non-standard files may only be partially indexed or skipped
• ppt legacy extraction is intentionally stricter than modern pptx

Good fit

• contract folders
• report archives
• bid and compliance materials
• policy and process folders
• project delivery folders
• mixed Office and PDF knowledge folders

Release

Current public version: v0.1.0