AI Case Manager

The AI case manager is your research partner—use it to analyze cases, draft documents, and understand complex legal concepts. Unlike ChatGPT, it sees your entire case context.

Important Reminder

AI can make mistakes, including citing non-existent cases. Always verify legal citations and consult with a licensed attorney before filing any documents.

Basic Usage

Opening the Chat

Access the AI assistant in several ways:

  • Click the Void icon in the left sidebar
  • Press Ctrl+L to focus the chat input
  • Press Ctrl+Shift+L to open the chat with your current selection

Sending Messages

  • Type your message in the input field
  • Press Enter to send (or Shift+Enter for a new line)
  • Wait for the AI to respond—longer responses may take a few seconds

Adding Context

Help the AI understand your needs by providing context:

  • @-mention files — Type @ and select files to include in your message
  • Select text first — Select text in the editor, then press Ctrl+Shift+L
  • Describe your case — The more context you provide, the better the response

Chat Modes

SafeAppeals offers three chat modes tailored for legal work:

Drafting Mode

Interactive document creation with AI guidance and direct editing capabilities.

Features

  • • Direct document editing (read, edit, create files)
  • • RAG search for policies and workspace content
  • • Timeline management tools
  • • Web search for current information

Example prompt

"Help me draft an introduction paragraph for my appeal letter that emphasizes the medical evidence."

Research Mode

Deep policy and case document analysis with comprehensive search tools.

Features

  • • All search and analysis tools enabled
  • • Safe exploration (no destructive actions)
  • • Policy document analysis
  • • Case precedent discovery

Example prompt

"Research recent workers' compensation decisions involving repetitive strain injuries and summarize the key findings."

Case Manager Mode

Proactive case workflow management with full autonomous tool access.

Features

  • • Full access to all built-in tools
  • • Autonomous multi-step task execution
  • • File organization and management
  • • Complete case workflow automation

Example prompt

"Analyze all documents in my Evidence folder, create a timeline of events, and identify gaps in my evidence."

Changing Modes

  1. Look for the mode dropdown at the top of the chat sidebar
  2. Click to open the dropdown
  3. Select your desired mode
💡

Recommendation: Start with Research mode for exploration, switch to Drafting when writing documents, and use Case Manager for complex multi-step workflows where you trust the AI to take actions autonomously.

Quick Edit (Ctrl+K)

Quick Edit is a powerful feature for making inline changes to your documents:

  1. Select the text you want to modify in the editor
  2. Press Ctrl+K
  3. Describe the change you want (e.g., "make this more formal")
  4. Press Enter to apply
  5. Review the changes and accept or reject them

Quick Edit Ideas

Improve clarity

Rewrite this to be clearer and more concise

Add formality

Make this more formal and professional

Expand content

Add more detail and supporting points

Fix errors

Fix any grammar or spelling mistakes

Add citation

Add a proper legal citation for this statement

Summarize

Condense this into a brief summary

Best Practices for Legal Research

Effective Prompting

  • Be specific — Include jurisdiction, type of claim, and specific issues
  • Provide context — Share relevant case details when asking for advice
  • Ask for sources — Request citations and ask the AI to verify them
  • Iterate — Follow up with clarifying questions

Verification Checklist

Before using AI-generated content:

  • Verify all case citations exist in legal databases
  • Confirm statutes and regulations are current
  • Check that legal principles apply to your jurisdiction
  • Review for any factual errors or inconsistencies
  • Have a licensed attorney review before filing

Token Usage

Each message uses tokens from your credit balance. Token usage depends on:

  • Input tokens — Your message length plus any files/context you include
  • Output tokens — The length of the AI's response
  • Model choice — Different models have different token costs

View your current balance in the Dashboard.

Next Steps

Case ManagementAI Models