@triedotdev/mcp 1.0.4 → 1.0.6

package/README.md

@@ -1,51 +1,28 @@
-# Quick Start Guide
+# Trie
 
-Get Trie Agent up and running in minutes to start scanning and fixing AI-generated code.
+**Customizable Agent Parallelization for AI Code Review**
 
-## Prerequisites
+13 specialized agents scan your code for security, privacy, compliance, and bugs—all running in parallel. Create custom agents from any document.
 
-- **Node.js 18+** installed ([download](https://nodejs.org/))
-- **AI coding tool** (Cursor, Claude Code, or VS Code with MCP support)
+## Features
 
-## Installation
+- **13 Built-in Agents** - Security, Privacy, SOC 2, Legal, Architecture, DevOps, and more
+- **Parallel Execution** - All agents run simultaneously for fast scans
+- **🔥 YOLO Mode** - Autonomous auto-fixing as you code
+- **Custom Agents** - Create agents from PDFs, docs, or style guides
+- **No API Key Required** - Works with any MCP-compatible AI tool (Cursor, Claude Code, VS Code)
+- **Smart Triaging** - Only activates relevant agents based on code context
+- **Docker Support** - Run in containers for CI/CD or isolated environments
 
-### Step 1: Install Trie Agent
+## Quick Start
 
 ```bash
 npm install -g @triedotdev/mcp
 ```
 
-### Step 2: Configure Your AI Tool
+### Configure Cursor
 
-No API key required! Trie uses your AI tool's built-in Claude to do the analysis.
-
-## Configuration
-
-### For Claude Code
-
-Use the `claude` CLI to add the MCP server:
-
-```bash
-claude mcp add Trie -- npx @triedotdev/mcp
-```
-
-Or add to `~/.claude/settings.json`:
-
-```json
-{
-  "mcpServers": {
-    "Trie": {
-      "command": "npx",
-      "args": ["@triedotdev/mcp"]
-    }
-  }
-}
-```
-
-### For Cursor
-
-1. Open Cursor Settings → MCP Servers
-2. Add this configuration:
+Settings → MCP Servers → Add:
 
 ```json
 {
@@ -58,231 +35,159 @@ Or add to `~/.claude/settings.json`:
 }
 ```
 
-### For VS Code
-
-Add to your VS Code settings (`.vscode/settings.json` or user settings):
+### Configure Claude Code
 
-```json
-{
-  "mcp.servers": {
-    "Trie": {
-      "command": "npx",
-      "args": ["@triedotdev/mcp"]
-    }
-  }
-}
+```bash
+claude mcp add Trie -- npx @triedotdev/mcp
 ```
 
 ## Usage
 
-### Scan Your Code
+Once configured, just ask your AI assistant:
 
-Once configured, use Trie Agent in your AI coding tool:
-
-**In Cursor/Claude Code:**
 ```
-Scan this code with Trie Agent
+Scan this code with Trie
 ```
 
-**Or use the MCP tool directly:**
+Or use specific agents:
+
 ```
-Use trie_scan to analyze the current file
+Run trie_security on this file
+Run trie_soc2 to check compliance
 ```
 
-### Example Workflow
-
-1. **Generate code** with your AI assistant
-2. **Scan for issues:**
-   ```
-   Scan this with Trie Agent
-   ```
-3. **Review results** - Trie will show:
-   - Risk level and activated agents
-   - Critical issues requiring review
-   - Auto-fixable issues
-   - Plain-language explanations
-4. **Apply fixes:**
-   ```
-   Auto-fix the high-confidence issues
-   ```
-5. **Generate tests:**
-   ```
-   Generate tests for this code
-   ```
+## YOLO Mode 🔥
 
-## Available Tools
-
-- **`trie_scan`** - Scan code with intelligent agent selection
-- **`trie_fix`** - Apply high-confidence fixes automatically
-- **`trie_explain`** - Get plain-language explanations of code/issues
-- **`trie_test`** - Generate tests or check coverage
-- **`trie_watch`** - Watch mode for continuous scanning
-- **`trie_security`**, **`trie_privacy`**, **`trie_bugs`**, etc. - Run specific agents
-
-### Custom Agent Tools
+**Autonomous auto-fixing** - Trie watches your code and automatically fixes high-confidence issues as you code.
 
-- **`trie_create_agent`** - Create a custom agent from a document (PDF, TXT, MD)
-- **`trie_save_agent`** - Save agent config after extraction
-- **`trie_list_agents`** - List all registered agents (built-in and custom)
-
-## Custom Agents
-
-Create specialized code review agents from any document—style guides, compliance docs, framework best practices, etc.
-
-### Creating a Custom Agent
+### Via MCP
 
 ```
-Use trie_create_agent with filePath:"./react-handbook.pdf" agentName:"react-handbook"
+Use trie_watch with action:"start" yolo:true
 ```
 
-Or paste content directly:
-```
-Use trie_create_agent with agentName:"my-style-guide" documentContent:"[paste your style guide here]"
+### Via CLI
+
+```bash
+trie-yolo
 ```
 
-The tool will parse the document and prompt you to extract knowledge. Then use `trie_save_agent` to save it.
+YOLO mode will:
+- Watch for file changes
+- Scan changed files automatically  
+- Auto-fix high-confidence issues (>95% confidence)
+- Log all actions for review
 
-### Listing Custom Agents
+## Docker
 
-```
-Use trie_list_agents
-```
+Run Trie in a container for CI/CD or isolated environments.
 
-This shows all built-in and custom agents with their pattern counts and categories.
+```bash
+# Build
+docker build -t trie-agent .
 
-### Using a Custom Agent
+# YOLO mode (auto-fix)
+docker run -v $(pwd):/app trie-agent --yolo
 
-Custom agents automatically activate during `trie_scan` based on their activation rules. To force a specific custom agent:
+# Watch mode (scan only)
+docker run -v $(pwd):/app trie-agent
 
+# CI mode (one-shot scan)
+docker run -v $(pwd):/app trie-agent --once
 ```
-Use trie_scan with forceAgents:["react-handbook"]
-```
-
-### Removing a Custom Agent
 
-Custom agents are stored as JSON files in your project's `.trie/agents/` directory. To remove one:
+Or use Docker Compose:
 
 ```bash
-# List custom agents
-ls .trie/agents/
+docker-compose up
+```
+
+| Mode | Command | Description |
+|------|---------|-------------|
+| YOLO | `--yolo` | Auto-fix high-confidence issues |
+| Watch | (default) | Scan on file changes, no auto-fix |
+| CI | `--once` | One-shot scan, exit with error code if issues found |
+
+## Built-in Agents
+
+| Agent | Description |
+|-------|-------------|
+| `security` | Vulnerabilities, injection risks, hardcoded secrets |
+| `privacy` | GDPR, CCPA, PII handling, data protection |
+| `soc2` | SOC 2 Type II compliance: access controls, encryption, logging |
+| `legal` | Consent patterns, data retention, compliance |
+| `architecture` | SOLID principles, code organization, scalability |
+| `bugs` | Null safety, edge cases, async issues |
+| `types` | Type errors, missing annotations |
+| `devops` | Config issues, logging, environment variables |
+| `accessibility` | WCAG 2.1, keyboard nav, screen readers |
+| `ux` | User flow testing, error states |
+| `clean` | AI-generated code cleanup |
+| `test` | Test generation and coverage |
+| `comprehension` | Plain-language explanations |
 
-# Remove a specific agent
-rm .trie/agents/react-handbook.json
-```
+## Custom Agents
+
+Create specialized agents from any document:
 
-The agent will no longer be loaded on subsequent scans.
+```
+Use trie_create_agent with filePath:"./style-guide.pdf" agentName:"my-style"
+```
 
-## MCP Resources
+Custom agents automatically activate during scans based on their rules.
 
-Trie also exposes read-only MCP resources for querying agent info, cache stats, and scan reports:
+## Available Tools
 
-| Resource | Description |
-|----------|-------------|
-| `trie://agents` | List all available agents |
-| `trie://config` | Current configuration |
-| `trie://cache/stats` | Cache performance metrics |
-| `trie://signatures` | Vulnerability signature counts |
-| `trie://reports/{file}` | Access scan reports |
+| Tool | Description |
+|------|-------------|
+| `trie_scan` | Intelligent scan with automatic agent selection |
+| `trie_fix` | Apply high-confidence fixes |
+| `trie_explain` | Plain-language explanations |
+| `trie_watch` | Continuous scanning mode |
+| `trie_create_agent` | Create custom agent from document |
+| `trie_list_agents` | List all available agents |
 
-See [MCP_USAGE.md](MCP_USAGE.md) for detailed resource documentation.
+Plus individual agent tools: `trie_security`, `trie_privacy`, `trie_soc2`, `trie_bugs`, etc.
 
 ## Example Output
 
 ```
-🔍 TRIE AGENT SCAN
+TRIE AGENT SCAN
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
 Smart Triaging:
 ✓ Risk Level: high
-✓ Reason: High-risk authentication feature
-✓ Agents activated: security, privacy, legal, architecture, typecheck
-✓ Execution time: 8.2s
-
-Results:
-Score: 72/100 (needs work)
-
-🔴 2 Critical Issues (require your review)
-🟡 3 Serious Issues (can auto-fix)
-🔵 2 Moderate Issues (can auto-fix)
-
-Critical Issues Preview:
-1. Password stored without hashing (auth/signup.ts:23)
-   Agent: Security | Confidence: 100%
-2. PII stored in plain text (models/user.ts:15)
-   Agent: Privacy | Confidence: 95%
+✓ Agents activated: security, soc2, privacy, architecture
+✓ Execution time: 3.2s
+
+Results: 72/100
+
+🔴 2 Critical Issues
+   • Hardcoded API key (CC6.1) - src/api.ts:15
+   • SQL injection risk - src/db.ts:42
+
+🟡 3 Serious Issues (auto-fixable)
+🔵 2 Moderate Issues
 ```
 
-## Configuration (Optional)
+## Configuration
 
-Create `.trie/config.json` in your project root to customize behavior:
+Create `.trie/config.json` to customize:
 
 ```json
 {
-  "version": "1.0",
-  "triaging": {
-    "enabled": true,
-    "riskThresholds": {
-      "critical": 70,
-      "high": 40,
-      "medium": 20
-    },
-    "autoFixConfidence": 0.95
-  },
   "agents": {
     "builtin": {
       "security": { "enabled": true },
-      "privacy": { "enabled": true },
-      "legal": { "enabled": true },
-      "design-engineer": { "enabled": true }
+      "soc2": { "enabled": true },
+      "privacy": { "enabled": true }
     }
+  },
+  "compliance": {
+    "standards": ["GDPR", "CCPA", "SOC2"]
   }
 }
 ```
+## License
 
-## Troubleshooting
-
-### Agent not found
-
-**Problem:** Can't find the Trie Agent executable
-
-**Solution:** Make sure you have Node.js 18+ installed and try:
-```bash
-npx @triedotdev/mcp --help
-```
-
-### No agents activating
-
-**Problem:** Scan runs but no agents are triggered
-
-**Solution:**
-- Check that your code matches patterns (auth, payments, UI, etc.)
-- Review the triaging logic in `.trie/config.json`
-- Try scanning a file with known issues (auth, payments, etc.)
-
-### MCP connection fails
-
-**Problem:** Can't connect to Trie Agent server
-
-**Solution:**
-1. Verify Node.js is installed: `node --version` (should be 18+)
-2. Check the path in MCP config is correct
-3. Restart your AI coding tool
-4. Check console/logs for error messages
-
-## Next Steps
-
-- **Try different code patterns** - Test with auth, payments, UI components
-- **Explore agents** - See which agents activate for different code types
-- **Use auto-fix** - Let Trie fix high-confidence issues automatically
-- **Generate tests** - Create comprehensive test suites
-- **Add custom agents** - Extend Trie with your own review logic
-
-## Support
-
-- 📚 **Documentation**: [trie.dev/docs](https://trie.dev/docs)
-- 🐛 **Issues**: [GitHub Issues](https://github.com/Trie-OS/mcp-agent/issues)
-- 💬 **Community**: [Discord](https://discord.gg/trie-ai)
-
----
-
-**Ready to ship safer code?** Start scanning your AI-generated code now! 🚀
+MIT