Workspace SOP logo

Workspace SOP

Community
wentorai
Workspace SOP

Workspace file management, git versioning, directory conventions, and CLI execution safety rules for the Research-Claw workspace.

Overview

Publisherwentorai
RepositoryResearch-Claw
Skill nameWorkspace SOP
Stars
859
Forks
117
Bundled files
Instructions only
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 wentorai on GitHub. Read the source before you install it.

Installation

Install the Workspace SOP 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/wentorai/Research-Claw/tree/main/skills/workspace-sop
  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/wentorai/Research-Claw.git /tmp/Research-Claw
mkdir -p .claude/skills
cp -r /tmp/Research-Claw/skills/workspace-sop .claude/skills/wentorai-workspace-sop
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Workspace SOP 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 Workspace SOP 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 Workspace SOP 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.

Workspace Operations SOP

Architecture

The workspace is a local Git repository, auto-initialized on first use. Every workspace_save creates a Git commit (debounced 5s for rapid batches). Files over 10 MB are auto-added to .gitignore. The repo is local-only — it never pushes to any remote.

The dashboard shows file tree, recent commits, and file previews. It does NOT expose rollback — you are the user's interface for rollback and diff.

Directory Conventions

Two top-level user directories (system .ResearchClaw/ is hidden from the dashboard):

sources/       ← user uploads, imported PDFs, datasets, references
sources/chat/  ← raw files the user dropped into the chat composer
outputs/       ← agent-generated drafts, figures, exports, reports
.ResearchClaw/ ← system prompt files (hidden)

Optional subdirs (created on demand, not required): sources/papers/, sources/data/, sources/references/ outputs/drafts/, outputs/figures/, outputs/exports/, outputs/reports/

sources/ = user input. outputs/ = agent-generated files. sources/chat/ holds timestamped raw drops from the chat box (images land in sources/chat/images/) — when the user asks you to organize files, file them from here into sources/papers/ etc.

Commit Message Prefixes

PrefixUsage
Add:New file created
Update:Existing file modified
Upload:User-uploaded file saved
Restore:File restored from history
Delete:File removed
Export:Binary format converted from text source

Tool Chain (11 tools)

ToolPurpose
workspace_saveWrite text file + auto-commit → Dashboard projects the file fact
workspace_readRead file contents
workspace_listList files (glob filter + git status)
workspace_diffShow changes (no range = uncommitted vs HEAD)
workspace_historyList commits for a file or path
workspace_restoreCheckout historical version + commit as Restore:
workspace_moveRename/move file + commit
workspace_exportConvert text → binary format (md→docx/pdf, csv→xlsx)
workspace_deleteDelete a file + auto-commit (requires confirm=true) → recoverable via workspace_restore
workspace_appendAppend content to an existing file (or create) → automatic file projection
workspace_downloadDownload URL → save binary to workspace → automatic file projection

Version Control Operations

  • Undo / rollback: workspace_history → present commits → user selects → workspace_restore → report the restored path in normal text.
  • Diff: workspace_diff (default: uncommitted vs HEAD).
  • Proactive: Mention git history after overwrites. Note workspace_restore on delete.
  • Delete: workspace_delete (requires confirm=true). Always mention workspace_restore as recovery option.

Write Discipline

  • Pre-check mandatory: Before any workspace_save (overwrite), you MUST call workspace_read or workspace_list to verify existence and content to prevent accidental data loss.
  • Append vs. Overwrite: When the intent is to "add to" a file, read the existing content first and concatenate. Never overwrite a multi-section file with only the new snippet.
  • Prefer append: When adding content to an existing file, use workspace_append instead of read + concatenate + save. It prevents accidental overwrites and uses fewer tokens.
  • Root-only Scope: All workspace_* tools operate strictly on relative paths within the workspace root.
  • Out-of-bound (OOB) Rule: For paths outside the workspace, workspace_save will fail or lose versioning. Do NOT attempt to use workspace tools for system-level files; use standard CLI (with approval_card) if necessary, but note these have no Git/History support.

Binary Format Exports

workspace_save writes UTF-8 text only. For Word/Excel/PDF output:

  1. Save content as Markdown/CSV: workspace_save("outputs/drafts/paper.md", content)
  2. Convert: workspace_export({ source: "outputs/drafts/paper.md", format: "docx" })
Source→ TargetEngine
.md .txt.docxpandoc
.md .txt.pdfpandoc + xelatex (CJK supported)
.csv .json.xlsxPython pandas + openpyxl

Output defaults to outputs/exports/{name}.{format}. Override with output param.

NEVER call workspace_save with a binary extension (.docx, .xlsx, .pdf, .pptx, etc.) — the tool will reject the write and guide you to workspace_export.

Native Install Dependencies

Docker images ship with all dependencies pre-installed. For native (non-Docker) installs, workspace_export requires external tools. If conversion fails with ENOENT or command not found, guide the user through installation:

macOS (Homebrew):

bash
# Install Homebrew first if missing:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

# Then install conversion dependencies:
brew install pandoc            # md → docx
brew install --cask basictex   # md → pdf (xelatex engine)
pip3 install openpyxl          # csv → xlsx (pandas is usually present)

Windows: Download pandoc installer from https://pandoc.org/installing.html, install MiKTeX for PDF support.

Note: .pptx generation is NOT supported by workspace_export. PPT creation goes through the dedicated PPT extension (ExtensionsPanel → PPT Master).

Download & Import

Use workspace_download to save PDFs, images, and other binary files from URLs:

workspace_download({ url: "https://arxiv.org/pdf/2301.00001", path: "sources/papers/paper.pdf" })

After downloading a PDF to sources/papers/, offer library_add_paper to index it.

CLI Execution Safety

Safe without approval (no approval_card needed):

wc, du, grep, find, pandoc, pdftotext, python3, xelatex, jq

Requires approval_card:

pip install, brew install, curl, wget, any operation outside workspace.

Cross-Module Triggers

  • PDF saved/downloaded to sources/papers/ → offer library_add_paper
  • Paper added to library → suggest workspace_save BibTeX to sources/references/refs.bib via workspace_append
  • Code file created → suggest task_create to track the work
  • Analysis output generated → report its relative path + task_complete if linked

Related Research-Plugins Skills

  • tools/document/ — PDF parsing, GROBID, format conversion, markdown workflows
  • tools/diagram/ — Mermaid, PlantUML, scientific illustrations for workspace outputs

Frequently asked questions

What does the Workspace SOP AI skill do?

Workspace file management, git versioning, directory conventions, and CLI execution safety rules for the Research-Claw workspace.

Why use Workspace SOP on TypingMind?

Because you install it once and use it with any model. Workspace SOP 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 Workspace SOP in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/wentorai/Research-Claw/tree/main/skills/workspace-sop. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Workspace SOP?

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 Workspace SOP?

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

Is the Workspace SOP AI skill free?

It is published on GitHub by wentorai. Check the repository for licensing terms. 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 👇