- Create comprehensive Cowork docs (23 files in cowork/) - 6 guides: overview, getting started, capabilities, security, troubleshooting - 60+ ready-to-use prompts across 4 categories - 5 step-by-step workflows - Reference materials: cheatsheet, FAQ, comparison, glossary - Integrate Perplexity research (P0/P1/P2) - Exact error messages with solutions (VPN, Chrome host, context limits) - Competitive analysis (vs Copilot/Gemini/ChatGPT/Apple Intelligence) - Enterprise validation (TELUS, Rakuten, Zapier stats) - OCR accuracy benchmarks (97% field, 63% line-item) - Token budget planning per task type - Document critical limitations - VPN incompatibility (#1 community issue) - Context limit reality (165K vs 200K theoretical) - Platform constraints (macOS only) - Usage limits and pricing (Pro $20, Max $100-200) - Update central files - README.md: detailed Cowork section with tables - VERSION: 3.9.6 → 3.9.7 - machine-readable/reference.yaml: add cowork_reference entry - machine-readable/cowork-reference.yaml: new LLM-optimized index (~1.5K tokens) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
4.4 KiB
Cowork Glossary
Terminology specific to Claude Cowork and related concepts
A
Agentic
Describes AI systems that can autonomously plan and execute multi-step tasks with minimal human intervention. Cowork is "agentic" because it analyzes your request, creates a plan, and executes it (with your approval).
Approval Gate
The checkpoint where Cowork shows its planned actions and waits for your confirmation before executing. Critical security measure—always review before approving.
C
Claude Code
Anthropic's CLI tool for developers. Shares architecture with Cowork but provides full shell access and code execution. Terminal-based interface.
Claude Desktop
The macOS application that hosts Cowork. Different from the web interface (claude.ai).
Context Limit
The maximum amount of text/data Claude can process in a single session (~200K tokens). When exceeded, tasks may fail or produce incomplete results.
Context Window
The "memory" space where Claude holds your conversation and file contents during a session. Measured in tokens.
Cowork
Claude's agentic desktop feature for knowledge workers. Manipulates files without code execution.
E
Execution Plan
The detailed list of actions Cowork proposes before starting work. Shows what files will be affected and how. Review this carefully.
Extended Thinking
Claude's ability to reason through complex problems step-by-step. Shared capability between Claude Code and Cowork.
F
Folder Sandbox
The restricted area where Cowork can operate. You grant access to specific folders; Cowork cannot access anything outside this boundary.
K
Knowledge Worker
Non-technical professionals who work primarily with information and documents (project managers, analysts, writers, consultants). Cowork's target audience.
L
Local-First
Design principle where data and processing stay on your computer rather than in the cloud. Cowork accesses local files only.
M
Max Tier
The highest Claude subscription level ($100-200/month, with 5x or 20x usage multipliers). Recommended for heavy Cowork usage.
Pro Tier
The standard Claude subscription level ($20/month). Now includes Cowork access, but with tighter usage limits (~1-1.5 hours intensive use before quota resets).
Multi-Step Task
An operation requiring multiple sequential actions. Example: "read files → analyze → create report → organize output." Cowork excels at these.
O
OCR (Optical Character Recognition)
Technology that extracts text from images. Cowork uses OCR to read receipts, screenshots, and scanned documents.
Orchestrator
The main Cowork agent that receives your request, creates the plan, coordinates sub-agents, and assembles results.
P
Prompt Injection
A security attack where malicious instructions are hidden in files, attempting to manipulate AI behavior. Mitigation: only process trusted files.
Projects
Claude's conversation interface on claude.ai. Allows document uploads but no local file access or creation.
R
Research Preview
Anthropic's term for early-access features that aren't production-ready. Cowork is in research preview (January 2026). Expect bugs.
S
Sub-Agent
Specialized workers spawned by the orchestrator to handle specific parts of a task. Each sub-agent has fresh context and can work in parallel with others.
Sandbox
See Folder Sandbox.
T
Token
The unit Claude uses to measure text. Roughly 4 characters or 0.75 words. Used to calculate context limits.
W
Workspace
The dedicated folder structure for Cowork operations. Best practice: ~/Cowork-Workspace/ with input/ and output/ subfolders.
Common Acronyms
| Acronym | Meaning |
|---|---|
| CLI | Command Line Interface |
| OCR | Optical Character Recognition |
| Portable Document Format | |
| API | Application Programming Interface |
| SSO | Single Sign-On |
Related Terms from Claude Code
| Term | In Code Context | In Cowork Context |
|---|---|---|
| CLAUDE.md | Project context file | Can use for shared team context |
| Sub-agents | Task-specific workers | Same concept, file-focused |
| Hooks | Event handlers | Not available in Cowork |
| MCP | Model Context Protocol | Not exposed in Cowork |