Office Document Review logo

Office Document Review

Community
AtlasOmnia
office-document-review

office-document-review — Review and proofread office documents in OneDrive or local file trees, including batch folder discovery, document text extraction, and spelling/grammar cleanup workflows.

Overview

PublisherAtlasOmnia
Repositorydonna-starter
Skill nameoffice-document-review
Stars
131
Forks
9
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by AtlasOmnia on GitHub. Read the source before you install it.

Installation

Install the Office Document Review AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/AtlasOmnia/donna-starter/tree/main/skills/productivity/office-document-review
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/AtlasOmnia/donna-starter.git /tmp/donna-starter
mkdir -p .claude/skills
cp -r /tmp/donna-starter/skills/productivity/office-document-review .claude/skills/office-document-review
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Office Document Review in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Office Document Review on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Office Document Review is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Office Document Review

Use this skill when the user wants you to inspect a OneDrive/SharePoint-style folder tree, locate a subfolder by name/date, and proofread all documents in it for spelling and grammar.

This is a workflow skill, not a single-app skill: it covers local file-system browsing, document text extraction, and batch proofreading notes.

Triggers

  • "go to the OneDrive folder"
  • "look in folder X and then folder Y"
  • "proofread all those documents"
  • "check for spelling/grammar errors in the docs"
  • "review the files in this folder"

Core workflow

  1. Resolve the real path first
  • Prefer a filesystem path over guessing from the user’s wording.
  • Folder names in casual speech may not match the actual on-disk name exactly.
  • If the exact folder cannot be found, list the visible folder names and state what is missing before asking for clarification.
  1. Search the tree recursively
  • Use the live file tree to locate the target folder and confirm the actual path.
  • When the user gives a date-like folder name, search for variant spellings and punctuation (6-2-26, 6.2.26, 6_2_26, 6-2-2026) before assuming it does not exist.
  1. Inventory the documents
  • Identify all files in the target folder.
  • Group by extension and prioritize editable office docs (.docx, .doc, .odt, .rtf, .txt) before PDFs or spreadsheets.
  1. Extract readable text
  • For Word/Office files, extract the text in a way that preserves headings and paragraphs as much as possible.
  • If the file-reading tool rejects .docx as binary, do not classify the document as unreadable and do not repeat the same call across the batch. A DOCX is a ZIP package: validate it with Python zipfile, then use python-docx when available or extract word/document.xml and convert paragraph/table XML to bounded text. Record package validity, paragraph/table counts, and extraction method. Use native Word or PDF rendering separately when layout—not just text—is under review.
  • When validating generated fixtures, compare extracted text against the fixture specification or authoritative .txt source before opening Word. A valid ZIP alone does not prove the fixture contains the intended content.
  • For scanned or image-based files, use OCR or an extraction tool first.
  • If a file is not meaningfully text-reviewable after at least one appropriate format-specific fallback, say so explicitly instead of pretending to proofread it.
  1. Proofread for two passes
  • Pass 1: spelling, grammar, punctuation, capitalization.
  • Pass 2: obvious clarity issues, duplicated words, inconsistent dates/names, and broken formatting that would change meaning.
  • Keep business wording intact unless the user asked for rewriting.
  1. Report findings file-by-file
  • Summarize each document with:
  • file name
  • issue count
  • short list of fixes or concerns
  • If no issues are found, say that clearly.
  1. Do not silently modify files unless asked
  • The default action is review/report.
  • If the user wants edits, confirm the target files and whether they want tracked wording changes or a clean rewrite.

Pitfalls

  • Casual folder names may be approximate; always verify the actual folder names seen in the tree.
  • A folder search that returns the parent tree but not the exact folder is not a success; say what was found and what was not.
  • If the user says they downloaded the files, re-inventory the local download folder first; do not assume the original folder contents or the original extension set survived the move.
  • Do not claim “everything is fine” for a batch until each file was actually inspected.
  • If a document is large, inspect enough of it to make a real proofreading judgment rather than sampling one paragraph and extrapolating.
  • Keep the user informed if the folder path is ambiguous instead of asking generic questions.

Verification

Before closing the task, confirm:

  • the folder path was resolved,
  • every target document was identified,
  • each document was reviewed or explicitly excluded with a reason,
  • and the final report separates true findings from non-reviewable files.

Session note file

Practical notes

  • Always verify the exact business OneDrive root before drilling into a dated subfolder.
  • If the user says "OneDrive" in a work/document-review context and does not qualify which account, default to the work tenant first (OneDrive - <work tenant>) rather than the personal OneDrive. Keep the two trees mentally separate and say which root you are using when ambiguity could matter.
  • If folder recursion on a synced OneDrive tree is unstable, list the parent first and narrow the search in smaller steps instead of forcing a full tree walk.
  • When direct recursion throws Resource deadlock avoided, try a shallow parent find/inventory pass and then inspect the exact subfolder file-by-file.
  • If the user says they downloaded the files, re-inventory the local download folder first; do not assume the original folder contents or extensions survived the move.
  • Keep the final review file-by-file so the user can see what was actually checked.

Session reference

  • When the user asks for "edits needed," separate true spelling/grammar fixes from OCR or extraction artifacts and label each file accordingly.

Frequently asked questions

What does the Office Document Review AI skill do?

office-document-review — Review and proofread office documents in OneDrive or local file trees, including batch folder discovery, document text extraction, and spelling/grammar cleanup workflows.

Why use Office Document Review on TypingMind?

Because you install it once and use it with any model. Office Document Review is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Office Document Review in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/AtlasOmnia/donna-starter/tree/main/skills/productivity/office-document-review. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Office Document Review?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Office Document Review?

As many as you like. As long as a model supports skills, you can use Office Document Review with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Office Document Review AI skill free?

Yes. It is published on GitHub by AtlasOmnia under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇