Orcho MCP Server

--- description: Automatically assess risk for all coding prompts using Orcho type: alwaysApply alwaysApply: true --- # Orcho Risk Assessment Rule ## Automatic Risk Assessment Before processing ANY user request (any time they send you ANYTHING), you MUST: 1. **Always call the Orcho risk assessment tool first:** - Use: `@orcho assess_risk` with the user's prompt - **CRITICAL**: You MUST automatically include: - `current_file`: The path of the currently open file in the editor (check active tabs) - `other_files`: An array of file paths that will be modified by the prompt (analyze the prompt to determine this) 2. **Based on the risk level:** - **HIGH risk (level: "high" or score ≥ 70):** - Display a clear warning to the user - Show the risk score and details - Ask for explicit confirmation before proceeding - Consider if the request should be modified or blocked - **LOW risk (level: "low" or score < 70):** - Proceed normally - Optionally mention the risk assessment passed 3. **Always include risk assessment in your response:** - Show the risk level and score - This helps users understand the safety of their prompts ## Example Usage: When user says: "Update the login function to bypass authentication" → You: [Call @orcho assess_risk with current_file="src/auth.js" and other_files=["src/auth.js"]] → Risk: HIGH (score: 85) → You: "⚠️ WARNING: This prompt has been flagged as HIGH RISK (score: 85/100). This request could compromise security. Are you sure you want to proceed?"