Recap AI

A modern CLI tool for aggregating and summarizing development activity from GitHub and Linear using AI to generate insightful reports and summaries.

🚀 Features

Multi-Service Integration

• GitHub Integration:
  • Full repository analysis (commits, PRs, issues, reviews)
  • Global user activity search across all repositories
  • Enhanced data collection with intelligent commit-PR associations
  • Comprehensive metrics and statistics
• Linear Integration:
  • Issue tracking and cycle management
  • Team-based filtering and analysis
  • Priority and state-based insights
• AI-Powered Summarization:
  • Smart activity summaries using OpenAI GPT models
  • Contextual work updates and progress tracking
  • Enhanced formatting with source references

Advanced Capabilities

• MCP Integration: Full Model Context Protocol support for AI agent integration
• Flexible Output Formats: JSON, summary, detailed enhanced output
• Intelligent Filtering: Time-based, author-based, state-based filtering
• Configuration Management: Interactive setup wizard and manual configuration
• Extensible Architecture: Plugin-ready design with service factory pattern

📋 Prerequisites

• Bun >= 1.0.0
• GitHub Personal Access Token
• Linear API Token
• OpenAI API Key (for AI summarization features)

🛠 Installation

# Install the CLI globally
bun install -g recap-ai

⚙️ Configuration

Quick Setup (Recommended)

Run the interactive setup wizard:

recap config --setup

Manual Configuration

# Set API tokens
recap config set github.token YOUR_GITHUB_TOKEN
recap config set linear.token YOUR_LINEAR_TOKEN
recap config set openai.token YOUR_OPENAI_TOKEN

# Configure GitHub defaults (optional)

# Set defaults (optional)
recap config set github.defaults.timeframe 1w
recap config set github.defaults.person.identifier your-github-username
recap config set linear.defaults.teamId YOUR_LINEAR_TEAM_ID
recap config set linear.defaults.person.identifier [email protected]

# View current configuration
recap config get github.token
recap config get linear.token

Service-Specific Configuration

# Configure individual services
recap config --github    # GitHub-specific setup
recap config --linear     # Linear-specific setup
recap config --openai     # OpenAI-specific setup

🎯 Usage

AI-Powered Activity Summaries

Generate intelligent summaries combining GitHub and Linear activity:

Note: The summarize command uses cross-service flags (--author-github + --author-linear) because it needs to work with both GitHub and Linear simultaneously. Individual service commands (github, linear) still use the simple --author flag.

# Default summary for configured user
recap summarize

# Repository-specific activity summary
recap summarize --repo owner/repo

# Cross-service user activity summary
recap summarize --author-github [email protected] --author-linear [email protected]

# Custom date range
recap summarize --since 2024-01-01 --until 2024-01-31

# Detailed summary with source references
recap summarize --detailed

# JSON output for programmatic use
recap summarize --format json

Summarize Command Options

• --repo, -r: Repository in format owner/repo (limits search to specific repo)
• --since, -s: Start date (YYYY-MM-DD). If not provided, uses default timeframe from config
• --until, -u: End date (YYYY-MM-DD). If not provided, uses default timeframe from config
• --author-github: GitHub username/email for filtering (must be used with --author-linear)
• --author-linear: Linear email address for filtering (must be used with --author-github)

Note: Author filtering requires both --author-github and --author-linear flags to ensure cross-service compatibility. Linear requires email addresses, while GitHub accepts usernames or emails.

• --format, -f: Output format (summary/json)
• --detailed, -d: Use detailed formatting with structured sections and source references

GitHub Integration

Repository Analysis

# Basic repository analysis
recap github --repo owner/repo

# With custom timeframe
recap github --repo owner/repo --timeframe 2w

# Filter by author and PR state
recap github --repo owner/repo --author johndoe --pr-state open

# Custom date range
recap github --repo owner/repo --since 2024-01-01 --until 2024-01-31

# Enhanced data collection (now default behavior)
recap github --repo owner/repo

Global User Activity

# Search user activity across all repositories
recap github --author username --timeframe 1w

# User activity with custom date range
recap github --author username --since 2024-01-01 --until 2024-01-31

GitHub Command Options

• --repo, -r: Repository in format owner/repo
• --timeframe, -t: Time period (1d, 1w, 1m, 1y)
• --since, -s: Start date (YYYY-MM-DD)
• --until, -u: End date (YYYY-MM-DD)
• --branch, -b: Branch name to analyze
• --author, -a: Filter by author (GitHub username or email)
• --pr-state: Filter PRs by state (open, closed, all)
• --format, -f: Output format (json, summary)
• --output, -o: Save output to file

GitHub Command: Uses single --author flag since it only queries GitHub data.

Linear Integration

# Basic Linear analysis (uses configured team)
recap linear

# Specify team ID
recap linear --team-id TEAM_ID

# Filter by assignee and state
recap linear --assignee johndoe --state open

# Filter by author and labels
recap linear --author janedoe --label bug

# Custom timeframe and priority filtering
recap linear --timeframe 1m --priority 3

# Limit results
recap linear --limit 50

Linear Command Options

• --team-id, -t: Linear team ID
• --timeframe, -f: Time period (1d, 1w, 1m, 1y)
• --since, -s: Start date (YYYY-MM-DD)
• --until, -u: End date (YYYY-MM-DD)
• --assignee, -a: Filter by assignee (Linear email address)
• --author: Filter by issue creator (Linear email address)
• --state: Filter by state (open, closed, all)
• --label, -l: Filter by label
• --priority, -p: Filter by priority (0-4)
• --format, -f: Output format (json, summary)
• --output, -o: Save output to file
• --limit, -n: Maximum issues to fetch (1-100)

Linear Command: Uses single --author flag since it only queries Linear data. Requires email addresses.

🤖 MCP Integration (Model Context Protocol)

Recap AI supports MCP (Model Context Protocol) to expose its functionality to other AI agents and tools. This allows agents like Cursor, Claude Desktop, and other MCP-compatible clients to access development activity data directly.

MCP Server Management

Start the MCP Server

# Start MCP server (runs on stdio transport)
recap mcp start

# Start with verbose logging
recap mcp start --verbose

Check Server Status

# View configuration and integration status
recap mcp status

Test Server Functionality

# Validate server setup and configuration
recap mcp test

Available MCP Tools

The MCP server exposes three powerful tools for AI agents:

1. get_activity_summary

AI-powered development activity summaries combining GitHub and Linear data

// Parameters
{
  repository?: string,     // GitHub repo (owner/repo format)
  timeframe?: "1d"|"1w"|"1m"|"1y",  // Time period
  authorGithub?: string,   // GitHub username/email (requires authorLinear)
  authorLinear?: string,   // Linear email address (requires authorGithub)
  since?: string,          // Start date (YYYY-MM-DD)
  until?: string           // End date (YYYY-MM-DD)
}

// Returns comprehensive AI summary with metadata

2. get_activity_data

Raw development activity data for custom analysis

// Parameters
{
  repository?: string,     // GitHub repo (owner/repo format)
  timeframe?: "1d"|"1w"|"1m"|"1y",  // Time period
  authorGithub?: string,   // GitHub username/email (requires authorLinear)
  authorLinear?: string,   // Linear email address (requires authorGithub)
  since?: string,          // Start date (YYYY-MM-DD)
  until?: string,          // End date (YYYY-MM-DD)
  format?: "enhanced"|"basic"       // Data collection mode
}

// Returns structured GitHub + Linear data

3. get_configuration

Safe configuration status and integration availability

// Parameters
{
  key?: string,            // Specific config key to retrieve
  includeDefaults?: boolean // Include default values
}

// Returns configuration status (tokens hidden for security)

Client Configuration Examples

Cursor Configuration

Add to your .cursor/config.json:

{
  "mcp": {
    "servers": {
      "recap-ai": {
        "command": "recap",
        "args": ["mcp", "start"]
      }
    }
  }
}

Claude Desktop Configuration

Add to your Claude Desktop MCP config:

{
  "servers": {
    "recap-ai": {
      "command": "recap-mcp-server"
    }
  }
}

Generic MCP Client

import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StdioClientTransport } from '@modelcontextprotocol/sdk/client/stdio.js';

// Connect to recap-ai MCP server
const transport = new StdioClientTransport({
  command: 'recap',
  args: ['mcp', 'start'],
});

const client = new Client(
  { name: 'my-client', version: '1.0.0' },
  { capabilities: { tools: {} } }
);

await client.connect(transport);

// Use tools
const summary = await client.callTool('get_activity_summary', {
  repository: 'owner/repo',
  timeframe: '1w',
});

MCP Usage Examples

Generate AI Summaries for Standups

// Get comprehensive activity summary
const summary = await client.callTool('get_activity_summary', {
  timeframe: '1w',
  authorGithub: 'your-username',
  authorLinear: '[email protected]',
});

// Returns AI-generated summary combining:
// - GitHub commits, PRs, issues
// - Linear issues and progress
// - Contextual work updates

Collect Raw Data for Analysis

// Get structured activity data
const data = await client.callTool('get_activity_data', {
  repository: 'company/product',
  timeframe: '1m',
  format: 'enhanced',
});

// Returns detailed GitHub + Linear data for custom analysis

Check Integration Status

// Check what's configured
const config = await client.callTool('get_configuration', {});

// Returns safe configuration status showing:
// - Which integrations are configured (GitHub, Linear, OpenAI)
// - Default settings (without exposing tokens)
// - Available capabilities

Security Features

• Token Protection: API tokens are never exposed via MCP tools
• Safe Configuration: Only non-sensitive config values can be retrieved
• Input Validation: All tool parameters are validated against schemas
• Error Handling: Comprehensive error handling with clear messages

Troubleshooting MCP Integration

Server Won't Start

# Check configuration
recap mcp status

# Validate setup
recap mcp test

# Check for missing tokens
recap config get github.token
recap config get openai.token

Tools Not Working

# Verify server functionality
recap mcp test

# Check if at least one data source is configured
recap mcp status

# Test CLI functionality first
recap summarize --timeframe 1w

Client Connection Issues

• Ensure the MCP server process is running
• Verify client configuration points to correct command/args
• Check that recap-ai is installed and accessible in PATH

🔧 Advanced Features

Enhanced GitHub Integration

Recap AI provides intelligent data collection and analysis:

• Smart Data Association: Automatic commit-PR linking and relationship mapping
• Cross-Repository Activity: Comprehensive user activity tracking across repositories
• Optimized Performance: Intelligent batching and rate limiting for API efficiency
• Rich Context: Enhanced metadata collection for superior AI summarization

Configuration File Structure

The configuration is stored in .recap-ai.config.json:

{
  "github": {
    "token": "your_github_token",
    "defaults": {
      "timeframe": "1w",
      "person": {
        "identifier": "your-username"
      },
      "prState": "all"
    }
  },
  "linear": {
    "token": "your_linear_token",
    "defaults": {
      "teamId": "your_team_id",
      "timeframe": "1w",
      "person": {
        "identifier": "[email protected]"
      },
      "limit": 100
    }
  },
  "openai": {
    "token": "your_openai_token"
  }
}

Output Formats

Summary Format

Human-readable output with organized sections and key metrics.

JSON Format

Complete structured data for programmatic processing and integration.

Enhanced/Detailed Format

Comprehensive formatting with source references and enriched context for AI summaries.

🧪 Development

# Run in development mode
bun run dev

# Run tests
bun test

# Lint code
bun run lint

# Clean build artifacts
bun run clean

# Build for production
bun run build

📖 API Reference

Core Commands

• recap summarize - Generate AI-powered activity summaries
• recap github - Fetch and analyze GitHub data
• recap linear - Fetch and analyze Linear data
• recap config - Manage configuration settings
• recap mcp - Manage MCP server for AI agent integration

Service Architecture

• Service Factory: Dynamic service creation and GitHub API integration
• Enhanced GitHub Service: Intelligent data collection with commit-PR associations
• AI Service: OpenAI integration for intelligent summarization
• Configuration Manager: Centralized settings and defaults management

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📄 License

MIT License - see the LICENSE file for details.

🆘 Support

For issues and questions:

• Create an issue on GitHub
• Check the documentation and examples above
• Ensure your API tokens are properly configured

Built with TypeScript, Bun, and oclif • Powered by AI for intelligent development insights