sub-agents-mcp 0.1.4 → 0.2.0

package/README.md

@@ -1,34 +1,48 @@
 # Sub-Agents MCP Server
 
-Let your AI assistant (Cursor, Claude) use specialized sub-agents for specific tasks. For example, create a "test-writer" agent that writes tests, or a "code-reviewer" agent that reviews your code.
+[![npm version](https://img.shields.io/npm/v/sub-agents-mcp.svg)](https://www.npmjs.com/package/sub-agents-mcp)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-## Why sub-agents-mcp?
+Bring Claude Code–style sub-agents to any MCP-compatible tool.
 
-While Claude Code has excellent built-in sub-agent functionality, it's exclusive to Claude Code. This MCP server brings that same powerful sub-agent pattern to **ANY LLM tool that supports MCP** - including Cursor, Windsurf, and others.
+This MCP server lets you define task-specific AI agents (like "test-writer" or "code-reviewer") in markdown files, and execute them via Cursor CLI or Claude Code CLI backends.
 
-**TL;DR**: Experience Claude Code's sub-agent workflow everywhere.
+## Why?
 
-→ [Read the full story](https://dev.to/shinpr/bringing-claude-codes-sub-agents-to-any-mcp-compatible-tool-1hb9)
+Claude Code offers powerful sub-agent workflows—but they're limited to its own environment. This MCP server makes that workflow portable, so any MCP-compatible tool (Cursor, Claude Desktop, Windsurf, etc.) can use the same agents.
 
-## Prerequisites
+**Concrete benefits:**
+- Define reusable agents once, use them across multiple tools
+- Share agent definitions within teams regardless of IDE choice
+- Leverage Cursor CLI or Claude Code CLI capabilities from any MCP client
 
-- Node.js 20 or higher
-- Cursor CLI or Claude Code installed
-- Basic terminal/command line knowledge
+→ [Read the full story](https://dev.to/shinpr/bringing-claude-codes-sub-agents-to-any-mcp-compatible-tool-1hb9)
 
-## Installation
+## Table of Contents
 
-```bash
-npx -y sub-agents-mcp
-```
+- [Prerequisites](#prerequisites)
+- [Quick Start](#quick-start)
+- [Usage Examples](#usage-examples)
+- [Agent Examples](#agent-examples)
+- [Configuration Reference](#configuration-reference)
+- [Session Management](#session-management)
+- [Troubleshooting](#troubleshooting)
+- [Design Philosophy](#design-philosophy)
+- [How It Works](#how-it-works)
+
+## Prerequisites
 
-This command will install and run the MCP server. No manual building or cloning required!
+- Node.js 20 or higher
+- One of these execution engines (they actually run the sub-agents):
+  - `cursor-agent` CLI (from Cursor)
+  - `claude` CLI (from Claude Code)
+- An MCP-compatible tool (Cursor IDE, Claude Desktop, Windsurf, etc.)
 
-## Quick Start (3 minutes)
+## Quick Start
 
-### Step 1: Create Your First Agent
+### 1. Create Your First Agent
 
-Create a folder for your agents and add a file `code-reviewer.md`:
+Create a folder for your agents and add `code-reviewer.md`:
 
 ```markdown
 # Code Reviewer
@@ -40,31 +54,36 @@ Focus on:
 - Checking code quality
 ```
 
-### Step 2: Setup Your AI Tool
+### 2. Install Your Execution Engine
+
+Pick one based on which tool you use:
 
-**For Cursor Users:**
+**For Cursor users:**
 ```bash
-# Install Cursor CLI
+# Install Cursor CLI (includes cursor-agent)
 curl https://cursor.com/install -fsS | bash
 
-# Login (Required!)
+# Authenticate (required before first use)
 cursor-agent login
 ```
 
-**For Claude Code Users:**
-
+**For Claude Code users:**
 ```bash
-# Option 1: NPM (requires Node.js 20+)
-npm install -g @anthropic-ai/claude-code
+# Option 1: Native install (recommended)
+curl -fsSL https://claude.ai/install.sh | bash
 
-# Option 2: Native install
-curl -fsSL claude.ai/install.sh | bash
+# Option 2: NPM (requires Node.js 18+)
+npm install -g @anthropic-ai/claude-code
 ```
 
-### Step 3: Configure MCP
+Note: Claude Code installs the `claude` CLI command.
+
+### 3. Configure MCP
 
-**For Cursor:** Edit `~/.cursor/mcp.json`
-**For Claude:** Edit `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS)
+Add this to your MCP configuration file:
+
+**Cursor:** `~/.cursor/mcp.json`
+**Claude Desktop:** `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS)
 
 ```json
 {
@@ -73,7 +92,7 @@ curl -fsSL claude.ai/install.sh | bash
       "command": "npx",
       "args": ["-y", "sub-agents-mcp"],
       "env": {
-        "AGENTS_DIR": "/path/to/your/agents-folder",  // ← Must be absolute path!
+        "AGENTS_DIR": "/absolute/path/to/your/agents-folder",
         "AGENT_TYPE": "cursor"  // or "claude"
       }
     }
@@ -81,59 +100,99 @@ curl -fsSL claude.ai/install.sh | bash
 }
 ```
 
-**Path examples:**
-- ✅ Good: `/Users/john/Documents/my-agents` (Mac/Linux)
-- ✅ Good: `C:\\Users\\john\\Documents\\my-agents` (Windows)
-- ❌ Bad: `./agents` or `~/agents` (relative paths don't work)
+**Important:** Use absolute paths only.
+- ✅ `/Users/john/Documents/my-agents` (Mac/Linux)
+- ✅ `C:\\Users\\john\\Documents\\my-agents` (Windows)
+- ❌ `./agents` or `~/agents` won't work
 
-**That's it!** Restart your IDE and start using agents.
+Restart your IDE and you're ready to go.
 
-## How to Use
+### 4. Fix "Permission Denied" Errors When Running Shell Commands
 
-Once configured, just tell your AI assistant to use your agents:
+Sub-agents may fail to execute shell commands with permission errors. This happens because sub-agents can't respond to interactive permission prompts.
 
-### Examples
+**Recommended approach:**
+
+1. Run your CLI tool directly with the task you want sub-agents to handle:
+   ```bash
+   # For Cursor users
+   cursor-agent
+
+   # For Claude Code users
+   claude
+   ```
+
+2. When prompted to allow commands (e.g., "Add Shell(cd), Shell(make) to allowlist?"), approve them
+
+3. This automatically updates your configuration file, and those commands will now work when invoked via MCP sub-agents
+
+**Manual configuration (alternative):**
+
+If you prefer to configure permissions manually, edit:
+- **Cursor**: `<project>/.cursor/cli.json` or `~/.cursor/cli-config.json`
+- **Claude Code**: `.claude/settings.json` or `.claude/settings.local.json`
+
+```json
+{
+  "permissions": {
+    "allow": [
+      "Shell(cd)",
+      "Shell(make)",
+      "Shell(git)"
+    ]
+  }
+}
+```
+
+Note: Agents often run commands as one-liners like `cd /path && make build`, so you need to allow all parts of the command.
+
+## Usage Examples
+
+Just tell your AI to use an agent:
 
-**Using a code reviewer:**
 ```
 "Use the code-reviewer agent to check my UserService class"
 ```
 
-**Using a test writer:**
 ```
 "Use the test-writer agent to create unit tests for the auth module"
 ```
 
-**Using a documentation writer:**
 ```
 "Use the doc-writer agent to add JSDoc comments to all public methods"
 ```
 
-Your AI will automatically invoke the specialized agent and return the results!
+Your AI automatically invokes the specialized agent and returns results.
+
+## Agent Examples
 
-## Common Agent Examples
+Each `.md` or `.txt` file in your agents folder becomes an agent. The filename becomes the agent name (e.g., `test-writer.md` → "test-writer").
 
-Here are some agents you might want to create:
+### Test Writer
 
-**`test-writer.md`** - Writes comprehensive unit tests
+**`test-writer.md`**
 ```markdown
 # Test Writer
-You are specialized in writing unit tests.
-- Write tests that cover edge cases
-- Follow the project's testing patterns
+You specialize in writing comprehensive unit tests.
+- Cover edge cases
+- Follow project testing patterns
 - Ensure good coverage
 ```
 
-**`sql-expert.md`** - Helps with database queries
+### SQL Expert
+
+**`sql-expert.md`**
 ```markdown
 # SQL Expert
-You are a database specialist.
-- Optimize queries for performance
+You're a database specialist who helps with queries.
+- Optimize for performance
 - Suggest proper indexes
 - Help with complex JOINs
 ```
 
-**`security-checker.md`** - Reviews code for security issues
+### Security Checker
+
+**`security-checker.md`**
 ```markdown
 # Security Checker
 You focus on finding security vulnerabilities.
@@ -142,95 +201,24 @@ You focus on finding security vulnerabilities.
 - Find potential data leaks
 ```
 
-## Configuration
+## Configuration Reference
 
-### Required Settings
+### Required Environment Variables
 
-**`AGENTS_DIR`** - Path to your agents folder
-- ⚠️ Must be an **absolute path**
-  - Mac/Linux: `/Users/john/my-agents`
-  - Windows: `C:\\Users\\john\\my-agents`
-- Create this folder before configuring MCP
+**`AGENTS_DIR`**
+Path to your agents folder. Must be absolute.
 
-**`AGENT_TYPE`** - Which AI tool you're using
-- Set to `"cursor"` for Cursor
-- Set to `"claude"` for Claude Code
+**`AGENT_TYPE`**
+Which execution engine to use:
+- `"cursor"` - uses `cursor-agent` CLI
+- `"claude"` - uses `claude` CLI
 
 ### Optional Settings
 
-**`EXECUTION_TIMEOUT_MS`** - How long agents can run (default: 5 minutes)
-- Increase for complex tasks like document review
-- Maximum: 10 minutes (600000ms)
-
-### Creating Agents
-
-Each `.md` or `.txt` file in your agents folder becomes an available agent.
-
-**File naming tips:**
-- Filename = agent name (e.g., `test-writer.md` → use as "test-writer")
-- Use hyphens or underscores, no spaces
-
-**Agent file structure:**
-```markdown
-# Agent Name
-Describe what this agent specializes in.
-List its key capabilities.
-```
-
-### Security Note
-
-Agents have access to your project directory. Only use agent definitions from trusted sources.
-
-## Troubleshooting
-
-### Cursor CLI Not Working
-
-**Symptoms:** Timeout errors, authentication failures, or "session expired" messages
-
-**Solutions:**
+**`EXECUTION_TIMEOUT_MS`**
+How long agents can run before timing out (default: 5 minutes, max: 10 minutes)
 
-1. **Authenticate with cursor-agent login**
-   ```bash
-   cursor-agent login
-   ```
-   This is the standard authentication method. Run this command before using the MCP server.
-
-2. **Check if cursor-agent is installed**
-   ```bash
-   which cursor-agent
-   ```
-   If not found, reinstall Cursor CLI.
-
-3. **Verify session status**
-   If you're still having issues, your session may have expired. Simply run `cursor-agent login` again.
-
-### Agent Not Found
-
-1. Verify `AGENTS_DIR` points to the correct directory
-2. Check file has `.md` or `.txt` extension
-3. Ensure filename contains only allowed characters
-
-### Other Execution Errors
-
-1. Verify `AGENT_TYPE` is set correctly (`cursor` or `claude`)
-2. Ensure the CLI tool is installed and accessible:
-   - For `cursor`: Ensure `cursor-agent` CLI is installed and authenticated
-   - For `claude`: Ensure Claude Code CLI is installed
-3. Check environment variables are properly set
-
-
-## How It Works
-
-Your AI assistant can invoke specialized agents through MCP:
-1. You ask your AI to use an agent (e.g., "Use the test-writer agent")
-2. The MCP server runs the specialized agent with your request
-3. Results come back to your main AI assistant
-
-## Additional Configuration Examples
-
-### Full Configuration Reference
-
-**For Cursor:** `~/.cursor/mcp.json`
+Example with timeout:
 ```json
 {
   "mcpServers": {
@@ -240,14 +228,32 @@ Your AI assistant can invoke specialized agents through MCP:
       "env": {
         "AGENTS_DIR": "/absolute/path/to/agents",
         "AGENT_TYPE": "cursor",
-        "EXECUTION_TIMEOUT_MS": "300000"  // Optional: 5 minutes default
+        "EXECUTION_TIMEOUT_MS": "600000"
       }
     }
   }
 }
 ```
 
-**For Claude:** `~/Library/Application Support/Claude/claude_desktop_config.json`
+### Security Note
+
+Agents have access to your project directory. Only use agent definitions from trusted sources.
+
+## Session Management
+
+Session management allows sub-agents to remember previous executions, which helps when you want agents to build on earlier work or maintain context across multiple calls.
+
+### Why Sessions Matter
+
+By default, each sub-agent execution starts with no context. With sessions enabled:
+- Agents can reference their earlier work
+- You get execution history for debugging
+- Related tasks share context
+
+### Enabling Sessions
+
+Add these environment variables to your MCP configuration:
+
 ```json
 {
   "mcpServers": {
@@ -256,18 +262,125 @@ Your AI assistant can invoke specialized agents through MCP:
       "args": ["-y", "sub-agents-mcp"],
       "env": {
         "AGENTS_DIR": "/absolute/path/to/agents",
-        "AGENT_TYPE": "claude",
-        "EXECUTION_TIMEOUT_MS": "300000"  // Optional
+        "AGENT_TYPE": "cursor",
+        "SESSION_ENABLED": "true",
+        "SESSION_DIR": "/absolute/path/to/session-storage",
+        "SESSION_RETENTION_DAYS": "7"
       }
     }
   }
 }
 ```
 
+**Configuration options:**
+
+- `SESSION_ENABLED` - Set to `"true"` to enable session management (default: `false`)
+- `SESSION_DIR` - Where to store session files (default: `.mcp-sessions` in the current working directory)
+- `SESSION_RETENTION_DAYS` - How long to keep session history in days (default: 7)
+
+**Security consideration:** Session files contain execution history and may include sensitive information. Use absolute paths for `SESSION_DIR`.
+
+### When to Use Sessions
+
+Sessions work well for:
+- **Iterative development**: "Based on your earlier findings, now fix the issues"
+- **Multi-step workflows**: Breaking complex tasks into smaller sub-agent calls
+- **Debugging**: Reviewing exactly what was executed and what results were returned
+
+Note that sessions require additional storage and processing overhead.
+
+### How Session Continuity Works
+
+When sessions are enabled, the MCP response includes a `session_id` field. To continue the same session, pass this ID back in the next request.
+
+**Important:** Your AI assistant must explicitly include the session_id in subsequent requests. While some assistants may do this automatically, it's not guaranteed. For reliable session continuity, add explicit instructions to your prompts or project rules.
+
+**Example prompt instruction:**
+```markdown
+When using sub-agents with sessions enabled, always include the session_id
+from the previous response in your next request to maintain context.
+```
+
+**Example project rule (e.g., `AGENTS.md`):**
+```markdown
+# Sub-Agent Session Guidelines
+
+When calling the same sub-agent multiple times:
+1. Extract the session_id from the MCP response
+2. Pass it as a parameter in subsequent calls
+3. This preserves context between executions
+```
+
+## Troubleshooting
+
+### Timeout errors or authentication failures
+
+**If using Cursor CLI:**
+Run `cursor-agent login` to authenticate. Sessions can expire, so just run this command again if you see auth errors.
+
+Verify installation:
+```bash
+which cursor-agent
+```
+
+**If using Claude Code:**
+Make sure the CLI is properly installed and accessible.
+
+### Agent not found
+
+Check that:
+- `AGENTS_DIR` points to the correct directory (use absolute path)
+- Your agent file has `.md` or `.txt` extension
+- The filename uses hyphens or underscores (no spaces)
+
+### Other execution errors
+
+1. Verify `AGENT_TYPE` is set correctly (`cursor` or `claude`)
+2. Ensure your chosen CLI tool is installed and accessible
+3. Double-check that all environment variables are set in the MCP config
+
+## Design Philosophy
+
+### Why Independent Contexts Matter
+
+Every sub-agent starts with a fresh context. This adds some startup overhead for each call, but it ensures that every task runs independently and without leftover state from previous runs.
+
+**Context Isolation**
+- Each agent only receives the information relevant to its task
+- No context leakage between runs
+- The main agent stays focused and lightweight
+
+**Accuracy and Reliability**
+- Sub-agents can specialize in a single goal without interference
+- Less risk of confusion from unrelated context
+- More consistent results in complex, multi-step workflows
+
+**Scalability**
+- Large tasks can be safely split into smaller sub-tasks
+- Each sub-agent operates within its own token limit
+- The main agent coordinates without hitting global context limits
+
+The startup overhead is an intentional trade-off: the system favors clarity and accuracy over raw execution speed.
+
+## How It Works
+
+This MCP server acts as a bridge between your AI tool and a supported execution engine (Cursor CLI or Claude Code CLI).
+
+**The flow:**
+
+1. You configure the MCP server in your client (Cursor, Claude Desktop, etc.)
+2. The client automatically launches `sub-agents-mcp` as a background process when it starts
+3. When your main AI assistant needs a sub-agent, it makes an MCP tool call
+4. The MCP server reads the agent definition (markdown file) and invokes the selected CLI (`cursor-agent` or `claude`)
+5. The execution engine runs the agent and streams results back through the MCP server
+6. Your main assistant receives the results and continues working
+
+This architecture lets any MCP-compatible tool benefit from specialized sub-agents, even if it doesn't have native support.
+
 ## License
 
 MIT
 
 ---
 
-*Enable AI-to-AI collaboration through Model Context Protocol*
+*AI-to-AI collaboration through Model Context Protocol*

package/dist/config/ServerConfig.d.ts

@@ -10,6 +10,9 @@
  * - AGENTS_DIR: Directory containing agent definition files (REQUIRED - must be absolute path)
  * - AGENT_TYPE: Type of agent to use ('cursor' | 'claude') (default: 'cursor')
  * - LOG_LEVEL: Log level for server operations (default: 'info')
+ * - SESSION_ENABLED: Enable session management functionality (default: false)
+ * - SESSION_DIR: Directory for storing session files (default: '.mcp-sessions')
+ * - SESSION_RETENTION_DAYS: Number of days to retain session files (default: 1)
  */
 export declare class ServerConfig {
     /** Server name identifier used for MCP registration */
@@ -24,6 +27,12 @@ export declare class ServerConfig {
     readonly logLevel: 'debug' | 'info' | 'warn' | 'error';
     /** Maximum execution timeout in milliseconds for agent execution */
     readonly executionTimeoutMs: number;
+    /** Enable session management functionality */
+    readonly sessionEnabled: boolean;
+    /** Directory for storing session files */
+    readonly sessionDir: string;
+    /** Number of days to retain session files before cleanup */
+    readonly sessionRetentionDays: number;
     /**
      * Creates a new ServerConfig instance by loading values from environment variables
      * or using default values.

package/dist/config/ServerConfig.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"ServerConfig.d.ts","sourceRoot":"","sources":["../../src/config/ServerConfig.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AACH,qBAAa,YAAY;IACvB,uDAAuD;IACvD,SAAgB,UAAU,EAAE,MAAM,CAAA;IAElC,6CAA6C;IAC7C,SAAgB,aAAa,EAAE,MAAM,CAAA;IAErC,gEAAgE;IAChE,SAAgB,SAAS,EAAE,MAAM,CAAA;IAEjC,yCAAyC;IACzC,SAAgB,SAAS,EAAE,QAAQ,GAAG,QAAQ,CAAA;IAE9C,sCAAsC;IACtC,SAAgB,QAAQ,EAAE,OAAO,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,CAAA;IAE7D,oEAAoE;IACpE,SAAgB,kBAAkB,EAAE,MAAM,CAAA;IAE1C;;;;OAIG;;CAkCJ"}
+{"version":3,"file":"ServerConfig.d.ts","sourceRoot":"","sources":["../../src/config/ServerConfig.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AACH,qBAAa,YAAY;IACvB,uDAAuD;IACvD,SAAgB,UAAU,EAAE,MAAM,CAAA;IAElC,6CAA6C;IAC7C,SAAgB,aAAa,EAAE,MAAM,CAAA;IAErC,gEAAgE;IAChE,SAAgB,SAAS,EAAE,MAAM,CAAA;IAEjC,yCAAyC;IACzC,SAAgB,SAAS,EAAE,QAAQ,GAAG,QAAQ,CAAA;IAE9C,sCAAsC;IACtC,SAAgB,QAAQ,EAAE,OAAO,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,CAAA;IAE7D,oEAAoE;IACpE,SAAgB,kBAAkB,EAAE,MAAM,CAAA;IAE1C,8CAA8C;IAC9C,SAAgB,cAAc,EAAE,OAAO,CAAA;IAEvC,0CAA0C;IAC1C,SAAgB,UAAU,EAAE,MAAM,CAAA;IAElC,4DAA4D;IAC5D,SAAgB,oBAAoB,EAAE,MAAM,CAAA;IAE5C;;;;OAIG;;CAiDJ"}

package/dist/config/ServerConfig.js

@@ -13,6 +13,9 @@ exports.ServerConfig = void 0;
  * - AGENTS_DIR: Directory containing agent definition files (REQUIRED - must be absolute path)
  * - AGENT_TYPE: Type of agent to use ('cursor' | 'claude') (default: 'cursor')
  * - LOG_LEVEL: Log level for server operations (default: 'info')
+ * - SESSION_ENABLED: Enable session management functionality (default: false)
+ * - SESSION_DIR: Directory for storing session files (default: '.mcp-sessions')
+ * - SESSION_RETENTION_DAYS: Number of days to retain session files (default: 1)
  */
 class ServerConfig {
     /**
@@ -48,6 +51,20 @@ class ServerConfig {
         else {
             this.executionTimeoutMs = 300000;
         }
+        // Session management configuration
+        // Only 'true' string enables session management, all other values are treated as false
+        this.sessionEnabled = process.env['SESSION_ENABLED'] === 'true';
+        this.sessionDir = process.env['SESSION_DIR'] || '.mcp-sessions';
+        // Parse SESSION_RETENTION_DAYS with validation (must be positive integer)
+        const retentionDaysEnv = process.env['SESSION_RETENTION_DAYS'];
+        if (retentionDaysEnv?.trim()) {
+            const parsedDays = Number.parseInt(retentionDaysEnv, 10);
+            // Use default (1 day) for invalid values (NaN, zero, or negative)
+            this.sessionRetentionDays = Number.isNaN(parsedDays) || parsedDays <= 0 ? 1 : parsedDays;
+        }
+        else {
+            this.sessionRetentionDays = 1;
+        }
     }
 }
 exports.ServerConfig = ServerConfig;

package/dist/config/ServerConfig.js.map

@@ -1 +1 @@
-{"version":3,"file":"ServerConfig.js","sourceRoot":"","sources":["../../src/config/ServerConfig.ts"],"names":[],"mappings":";;;AAAA;;;;;;;;;;;;GAYG;AACH,MAAa,YAAY;IAmBvB;;;;OAIG;IACH;QACE,IAAI,CAAC,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,aAAa,CAAC,IAAI,gBAAgB,CAAA;QAChE,IAAI,CAAC,aAAa,GAAG,OAAO,CAAC,GAAG,CAAC,gBAAgB,CAAC,IAAI,OAAO,CAAA;QAE7D,mDAAmD;QACnD,MAAM,SAAS,GAAG,OAAO,CAAC,GAAG,CAAC,YAAY,CAAC,CAAA;QAC3C,IAAI,CAAC,SAAS,EAAE,CAAC;YACf,MAAM,IAAI,KAAK,CACb,gDAAgD;gBAC9C,gEAAgE;gBAChE,gDAAgD;gBAChD,cAAc;gBACd,8DAA8D;gBAC9D,OAAO;gBACP,+BAA+B;gBAC/B,cAAc;gBACd,qDAAqD;gBACrD,KAAK,CACR,CAAA;QACH,CAAC;QACD,IAAI,CAAC,SAAS,GAAG,SAAS,CAAA;QAE1B,IAAI,CAAC,SAAS,GAAI,OAAO,CAAC,GAAG,CAAC,YAAY,CAAyB,IAAI,QAAQ,CAAA;QAC/E,IAAI,CAAC,QAAQ,GAAI,OAAO,CAAC,GAAG,CAAC,WAAW,CAAyC,IAAI,MAAM,CAAA;QAE3F,MAAM,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,sBAAsB,CAAC,CAAA;QACtD,IAAI,UAAU,EAAE,IAAI,EAAE,EAAE,CAAC;YACvB,MAAM,aAAa,GAAG,MAAM,CAAC,QAAQ,CAAC,UAAU,EAAE,EAAE,CAAC,CAAA;YACrD,IAAI,CAAC,kBAAkB,GAAG,MAAM,CAAC,KAAK,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,aAAa,CAAA;QAChF,CAAC;aAAM,CAAC;YACN,IAAI,CAAC,kBAAkB,GAAG,MAAM,CAAA;QAClC,CAAC;IACH,CAAC;CACF;AAzDD,oCAyDC"}
+{"version":3,"file":"ServerConfig.js","sourceRoot":"","sources":["../../src/config/ServerConfig.ts"],"names":[],"mappings":";;;AAAA;;;;;;;;;;;;;;;GAeG;AACH,MAAa,YAAY;IA4BvB;;;;OAIG;IACH;QACE,IAAI,CAAC,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,aAAa,CAAC,IAAI,gBAAgB,CAAA;QAChE,IAAI,CAAC,aAAa,GAAG,OAAO,CAAC,GAAG,CAAC,gBAAgB,CAAC,IAAI,OAAO,CAAA;QAE7D,mDAAmD;QACnD,MAAM,SAAS,GAAG,OAAO,CAAC,GAAG,CAAC,YAAY,CAAC,CAAA;QAC3C,IAAI,CAAC,SAAS,EAAE,CAAC;YACf,MAAM,IAAI,KAAK,CACb,gDAAgD;gBAC9C,gEAAgE;gBAChE,gDAAgD;gBAChD,cAAc;gBACd,8DAA8D;gBAC9D,OAAO;gBACP,+BAA+B;gBAC/B,cAAc;gBACd,qDAAqD;gBACrD,KAAK,CACR,CAAA;QACH,CAAC;QACD,IAAI,CAAC,SAAS,GAAG,SAAS,CAAA;QAE1B,IAAI,CAAC,SAAS,GAAI,OAAO,CAAC,GAAG,CAAC,YAAY,CAAyB,IAAI,QAAQ,CAAA;QAC/E,IAAI,CAAC,QAAQ,GAAI,OAAO,CAAC,GAAG,CAAC,WAAW,CAAyC,IAAI,MAAM,CAAA;QAE3F,MAAM,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,sBAAsB,CAAC,CAAA;QACtD,IAAI,UAAU,EAAE,IAAI,EAAE,EAAE,CAAC;YACvB,MAAM,aAAa,GAAG,MAAM,CAAC,QAAQ,CAAC,UAAU,EAAE,EAAE,CAAC,CAAA;YACrD,IAAI,CAAC,kBAAkB,GAAG,MAAM,CAAC,KAAK,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,aAAa,CAAA;QAChF,CAAC;aAAM,CAAC;YACN,IAAI,CAAC,kBAAkB,GAAG,MAAM,CAAA;QAClC,CAAC;QAED,mCAAmC;QACnC,uFAAuF;QACvF,IAAI,CAAC,cAAc,GAAG,OAAO,CAAC,GAAG,CAAC,iBAAiB,CAAC,KAAK,MAAM,CAAA;QAC/D,IAAI,CAAC,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,aAAa,CAAC,IAAI,eAAe,CAAA;QAE/D,0EAA0E;QAC1E,MAAM,gBAAgB,GAAG,OAAO,CAAC,GAAG,CAAC,wBAAwB,CAAC,CAAA;QAC9D,IAAI,gBAAgB,EAAE,IAAI,EAAE,EAAE,CAAC;YAC7B,MAAM,UAAU,GAAG,MAAM,CAAC,QAAQ,CAAC,gBAAgB,EAAE,EAAE,CAAC,CAAA;YACxD,kEAAkE;YAClE,IAAI,CAAC,oBAAoB,GAAG,MAAM,CAAC,KAAK,CAAC,UAAU,CAAC,IAAI,UAAU,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,UAAU,CAAA;QAC1F,CAAC;aAAM,CAAC;YACN,IAAI,CAAC,oBAAoB,GAAG,CAAC,CAAA;QAC/B,CAAC;IACH,CAAC;CACF;AAjFD,oCAiFC"}

package/dist/server/McpServer.d.ts

@@ -31,6 +31,7 @@ export declare class McpServer {
     private agentExecutor;
     private runAgentTool;
     private agentResources;
+    private sessionManager?;
     /**
      * Create a new MCP server instance
      * @param config Server configuration object

package/dist/server/McpServer.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"McpServer.d.ts","sourceRoot":"","sources":["../../src/server/McpServer.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;GAaG;AAeH,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,yBAAyB,CAAA;AAO3D;;GAEG;AACH,UAAU,UAAU;IAClB,IAAI,EAAE,MAAM,CAAA;IACZ,OAAO,EAAE,MAAM,CAAA;CAChB;AAOD;;GAEG;AACH,qBAAa,SAAS;IACpB,OAAO,CAAC,MAAM,CAAQ;IACtB,OAAO,CAAC,SAAS,CAAoC;IACrD,OAAO,CAAC,MAAM,CAAc;IAC5B,OAAO,CAAC,YAAY,CAAc;IAClC,OAAO,CAAC,aAAa,CAAe;IACpC,OAAO,CAAC,YAAY,CAAc;IAClC,OAAO,CAAC,cAAc,CAAgB;IAEtC;;;;OAIG;gBACS,MAAM,EAAE,YAAY;IA+ChC;;;;OAIG;IACH,OAAO,CAAC,cAAc;IAStB;;;;;OAKG;IACH,OAAO,CAAC,GAAG;IAqBX;;OAEG;IACH,OAAO,CAAC,cAAc;IAUtB;;OAEG;IACH,OAAO,CAAC,aAAa;IAuIrB;;;OAGG;IACH,aAAa,IAAI,UAAU;IAO3B;;;OAGG;IACH,YAAY,IAAI,OAAO;IAIvB;;;OAGG;IACH,OAAO,IAAI,OAAO;IAIlB;;;OAGG;IACG,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;IA2B5B;;;OAGG;IACG,SAAS,IAAI,OAAO,CAAC,KAAK,CAAC;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,OAAO,CAAA;KAAE,CAAC,CAAC;IAU9F;;;OAGG;IACG,aAAa,IAAI,OAAO,CAAC,KAAK,CAAC;QAAE,GAAG,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IASzF;;;;;OAKG;IACG,QAAQ,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,GAAG,OAAO,CAAC,OAAO,CAAC;IAOnE;;;;OAIG;IACG,YAAY,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,OAAO,CAAC;IAOjD;;;;OAIG;IACH,cAAc,IAAI;QAChB,UAAU,EAAE;YAAE,IAAI,EAAE,MAAM,CAAC;YAAC,OAAO,EAAE,MAAM,CAAA;SAAE,CAAA;QAC7C,cAAc,EAAE,GAAG,CAAC,MAAM,EAAE;YAAE,KAAK,EAAE,MAAM,CAAC;YAAC,SAAS,EAAE,MAAM,CAAC;YAAC,QAAQ,EAAE,IAAI,CAAA;SAAE,CAAC,CAAA;KAClF;IAOD;;OAEG;IACH,UAAU,IAAI,IAAI;IAIlB;;OAEG;IACG,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;CAuB7B"}
+{"version":3,"file":"McpServer.d.ts","sourceRoot":"","sources":["../../src/server/McpServer.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;GAaG;AAeH,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,yBAAyB,CAAA;AAQ3D;;GAEG;AACH,UAAU,UAAU;IAClB,IAAI,EAAE,MAAM,CAAA;IACZ,OAAO,EAAE,MAAM,CAAA;CAChB;AAOD;;GAEG;AACH,qBAAa,SAAS;IACpB,OAAO,CAAC,MAAM,CAAQ;IACtB,OAAO,CAAC,SAAS,CAAoC;IACrD,OAAO,CAAC,MAAM,CAAc;IAC5B,OAAO,CAAC,YAAY,CAAc;IAClC,OAAO,CAAC,aAAa,CAAe;IACpC,OAAO,CAAC,YAAY,CAAc;IAClC,OAAO,CAAC,cAAc,CAAgB;IACtC,OAAO,CAAC,cAAc,CAAC,CAAgB;IAEvC;;;;OAIG;gBACS,MAAM,EAAE,YAAY;IA+DhC;;;;OAIG;IACH,OAAO,CAAC,cAAc;IAStB;;;;;OAKG;IACH,OAAO,CAAC,GAAG;IAqBX;;OAEG;IACH,OAAO,CAAC,cAAc;IAUtB;;OAEG;IACH,OAAO,CAAC,aAAa;IAuIrB;;;OAGG;IACH,aAAa,IAAI,UAAU;IAO3B;;;OAGG;IACH,YAAY,IAAI,OAAO;IAIvB;;;OAGG;IACH,OAAO,IAAI,OAAO;IAIlB;;;OAGG;IACG,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;IA2B5B;;;OAGG;IACG,SAAS,IAAI,OAAO,CAAC,KAAK,CAAC;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,OAAO,CAAA;KAAE,CAAC,CAAC;IAU9F;;;OAGG;IACG,aAAa,IAAI,OAAO,CAAC,KAAK,CAAC;QAAE,GAAG,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IASzF;;;;;OAKG;IACG,QAAQ,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,GAAG,OAAO,CAAC,OAAO,CAAC;IAOnE;;;;OAIG;IACG,YAAY,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,OAAO,CAAC;IAOjD;;;;OAIG;IACH,cAAc,IAAI;QAChB,UAAU,EAAE;YAAE,IAAI,EAAE,MAAM,CAAC;YAAC,OAAO,EAAE,MAAM,CAAA;SAAE,CAAA;QAC7C,cAAc,EAAE,GAAG,CAAC,MAAM,EAAE;YAAE,KAAK,EAAE,MAAM,CAAC;YAAC,SAAS,EAAE,MAAM,CAAC;YAAC,QAAQ,EAAE,IAAI,CAAA;SAAE,CAAC,CAAA;KAClF;IAOD;;OAEG;IACH,UAAU,IAAI,IAAI;IAIlB;;OAEG;IACG,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;CAuB7B"}

package/dist/server/McpServer.js

@@ -21,6 +21,7 @@ const types_js_1 = require("@modelcontextprotocol/sdk/types.js");
 const AgentManager_1 = require("../agents/AgentManager");
 const AgentExecutor_1 = require("../execution/AgentExecutor");
 const AgentResources_1 = require("../resources/AgentResources");
+const SessionManager_1 = require("../session/SessionManager");
 const RunAgentTool_1 = require("../tools/RunAgentTool");
 const ErrorHandler_1 = require("../utils/ErrorHandler");
 const Logger_1 = require("../utils/Logger");
@@ -50,7 +51,22 @@ class McpServer {
         // Create logger with log level from config
         const executorLogger = new Logger_1.Logger(config.logLevel);
         this.agentExecutor = new AgentExecutor_1.AgentExecutor(executionConfig, executorLogger);
-        this.runAgentTool = new RunAgentTool_1.RunAgentTool(this.agentExecutor, this.agentManager);
+        // Initialize SessionManager if session management is enabled
+        if (config.sessionEnabled) {
+            this.sessionManager = new SessionManager_1.SessionManager({
+                enabled: config.sessionEnabled,
+                sessionDir: config.sessionDir,
+                retentionDays: config.sessionRetentionDays,
+            });
+            this.log('info', 'Session management enabled', {
+                sessionDir: config.sessionDir,
+                retentionDays: config.sessionRetentionDays,
+            });
+        }
+        else {
+            this.log('info', 'Session management disabled');
+        }
+        this.runAgentTool = new RunAgentTool_1.RunAgentTool(this.agentExecutor, this.agentManager, this.sessionManager);
         this.agentResources = new AgentResources_1.AgentResources(this.agentManager);
         // Initialize MCP server with capabilities
         this.server = new index_js_1.Server({
@@ -103,7 +119,7 @@ class McpServer {
                 service: 'mcp-server',
                 ...metadata,
             };
-            console.log(JSON.stringify(logEntry));
+            console.error(JSON.stringify(logEntry));
         }
     }
     /**