Opencode Analyzer

A comprehensive Python tool to analyze opencode logs and provide insights into development activity, patterns, and productivity metrics.

Features

• Session Analysis: Parse and reconstruct session timelines with full message history
• Tool Usage Statistics: Track tool frequency, distribution, and success rates
• Token Usage Analytics: Monitor input, output, and reasoning token consumption across models
• Content Theme Analysis: Identify themes, keywords, and patterns in development work
• Activity Patterns: Analyze hourly, daily, and weekly activity trends
• Progress Tracking: Track project evolution with trend analysis and milestone detection
• Error Analysis: Detect tool errors and debugging session patterns
• Search & Discovery: Full-text search across all prompts and responses
• Multiple Output Formats: Rich terminal tables, JSON, Markdown, and CSV
• Interactive CLI: Feature-rich command-line interface with 7 commands

Requirements

• Python 3.13+
• uv package manager

Installation

cd /path/to/opencode_analyzer
uv sync

Usage

Commands

# Show summary dashboard of all sessions
uv run opencode-analyzer overview

# List all sessions in reverse chronological order
uv run opencode-analyzer list

# Deep dive into a specific session
uv run opencode-analyzer analyze <session-id>

# Search through prompts and responses
uv run opencode-analyzer search "<query>"

# Generate time-based reports
uv run opencode-analyzer report --period today|week|month|all

# Show detailed statistics and metrics
uv run opencode-analyzer stats

# Output all JSON data for a session
uv run opencode-analyzer cat <session-id>

Global Options

All commands support these options:

Examples

# Overview with JSON output
uv run opencode-analyzer overview --format json

# Search with markdown output
uv run opencode-analyzer search "error handling" --format markdown

# Generate weekly report
uv run opencode-analyzer report --period week

# Analyze session with verbose output
uv run opencode-analyzer analyze abc123 --verbose

# Use custom data directory
uv run opencode-analyzer stats --opencode-dir /path/to/opencode

Project Structure

opencode_analyzer/
├── src/opencode_analyzer/
│   ├── __init__.py        # Package initialization
│   ├── cli.py             # Rich CLI interface (7 commands)
│   ├── data_parser.py     # Data parsing with caching
│   ├── analyzer.py        # Core analysis engines
│   ├── reporter.py        # Multi-format report generation
│   └── utils.py           # 40+ utility functions
├── tests/
│   ├── conftest.py        # Pytest fixtures
│   ├── unit/              # Component-specific tests
│   │   ├── test_data_parser.py
│   │   ├── test_analyzer.py
│   │   ├── test_cli.py
│   │   └── test_utils.py
│   └── integration/
│       └── test_end_to_end.py
├── pyproject.toml         # Project configuration
└── uv.lock                # Dependency lock file

Data Source

This tool analyzes opencode data stored in ~/.local/share/opencode/:

Analysis Capabilities

Session Metrics

• Total sessions, messages, and average metrics
• Model usage distribution
• Project correlation
• Time span and duration calculations

Tool Analysis

• Tool frequency and usage distribution
• Success and error rates per tool
• Sessions with specific tool patterns

Content Analysis

• Keyword extraction (top words and bigrams)
• Theme identification in prompts
• Response pattern analysis
• Reasoning content extraction

Activity Analysis

• Hourly, daily, weekly breakdowns
• Peak activity identification
• Session length distribution
• Productivity metrics

Progress Tracking

• Timeline reconstruction
• Cumulative metrics tracking
• Trend analysis (increasing/decreasing)
• Milestone identification

Dependencies

Core

• click - CLI framework
• rich - Terminal formatting and display
• orjson - Fast JSON parsing
• jsonschema - JSON validation
• python-dateutil - Date/time utilities
• pytz - Timezone support

Development

• pytest - Test framework
• pytest-cov - Coverage reporting
• pytest-mock - Mocking utilities

Development

# Run all tests
uv run pytest

# Run with coverage
uv run pytest --cov=src/opencode_analyzer

# Run specific test file
uv run pytest tests/unit/test_data_parser.py

# Run integration tests
uv run pytest tests/integration/

Performance

• Caching: LRU cache for message parsing (max 1000 items)
• Parallel Processing: ThreadPoolExecutor for session parsing (up to 4 workers)
• Fast JSON: Uses orjson for optimized JSON parsing
• Streaming: Handles large datasets efficiently

License

MIT