MCP Integration
The Model Context Protocol (MCP) is how your AI assistants interact with MoLOS. Think of it as a universal adapter that lets Claude, ChatGPT, or any AI tool see and use your data.
Why MCP Matters
| Problem | MCP Solution |
|---|---|
| AI forgets everything | MoLOS gives it permanent memory |
| Tools can't access data | MCP exposes your data as tools |
| No standard interface | MCP provides one protocol for all AIs |
| Security concerns | API keys let you control access |
With MCP enabled, your AI can:
- READ your tasks, notes, and projects
- CREATE new items when needed
- UPDATE status and progress
- STAY AWARE of your goals and priorities
Getting Started
Navigate to AI in main menu, then select MCP.
1. Create an API Key
Give your AI assistant access to MoLOS:
- Go to API Keys tab
- Click Create New Key
- Name it (e.g., "Claude Desktop")
- Choose permissions (which modules it can access)
- Click Generate and copy key immediately
Create secure access keys for your AI tools
Treat API keys like passwords. Use different keys for different tools, and revoke them immediately if compromised.
2. Connect Your AI Tool
Now use the key to connect:
| AI Tool | How to Connect |
|---|---|
| Claude Desktop | Settings → MCP → Add server → Enter key |
| ChatGPT | Use MCP-compatible plugin with your key |
| Custom Agents | Configure in your agent's MCP client |
Once connected, your AI can ask about your MoLOS data:
"What tasks do I have today?" → AI reads your Tasks module "Create a note about this project" → AI writes to your Markdown module "What's on my calendar?" → AI checks your Dashboard and Calendar
Create Custom Resources
Resources let you expose specific data or documents to your AI. Think of them as shortcuts for things your AI frequently needs.
Built-in Resources
MoLOS provides these automatically:
| Resource | URI | Description |
|---|---|---|
| molos://tasks | All your tasks, projects, and areas | |
| molos://markdown | Documents and quick notes | |
| molos://llm-council | Consultation history and personas |
Create Your Own
- Go to Resources tab
- Click Create Resource
- Give it a name and URI (e.g.,
molos://my-project) - Choose what it contains (static content or fetched from a URL)
- Save it
Use Cases:
- Project documentation that changes often
- Research notes you want AI to reference
- Custom data sources specific to your workflow
Reusable Prompts
Prompts save time by pre-defining common instructions for your AI.
Why Use Prompts?
| Problem | Prompt Solution |
|---|---|
| Repeating same instructions | Define once, reuse forever |
| Complex workflows | Break into named prompt templates |
| Team consistency | Everyone uses the same best prompts |
Create a Prompt
- Go to Prompts tab
- Click Create Prompt
- Name it (e.g., "Daily Standup")
- Add arguments (optional placeholders)
- Write prompt template with
{{argument}}syntax - Save
Example Prompt:
Review my tasks for today and suggest the top 3 priorities.
Context:
- Available time: {{time_available}}
- Energy level: {{energy_level}}
- Upcoming deadlines: {{check_deadlines}}
Format as a numbered list.
Now when you ask your AI for "daily standup," it uses this pre-configured prompt automatically.
Monitor Activity
The Activity tab shows every interaction between your AI tools and MoLOS.
| Why It Matters | What You See |
|---|---|
| Security | Log of every API call and data access |
| Debugging | Track down issues with AI connections |
| Audit trail | See what your AI has been doing |
Common Use Cases
For Individuals
| Goal | How MCP Helps |
|---|---|
| AI personal assistant | Claude Desktop with access to all your data |
| Task automation | AI creates and updates tasks automatically |
| Smart reminders | AI checks your schedule and suggests actions |
For Teams
| Goal | How MCP Helps |
|---|---|
| Shared knowledge base | Team members access documents via MCP |
| Project coordination | AI reads and updates shared tasks |
| Consistent workflows | Everyone uses the same prompt templates |
Troubleshooting
| Issue | Solution |
|---|---|
| AI can't connect? | Verify API key is correct and not expired |
| "Access denied" error? | Check key permissions include the module AI needs |
| Resource not found? | Verify URI matches molos://your-resource format |
| Want to revoke access? | Go to API Keys and delete the key immediately |
Learn More
- AI Chat Guide - Using the Architect Agent
- AI Tools Browser - Browse all available tools
- Core Modules - See what's available
Last Updated: March 21, 2026 | Version 1.1.0