🧠 DockAI Documentation

Welcome to the official DockAI documentation — an AI-powered CLI tool that analyzes Docker logs intelligently.

DockAI helps developers, DevOps engineers, and teams diagnose issues faster, summarize logs, and understand container behavior using Large Language Models (LLMs).

🚀 Getting Started

DockAI is available via PyPI and can be installed with:

pip install dockai

Once installed, you can analyze any running container logs:

dockai analyze <container-name> --since 15m --tail 3000

For performance metrics (CPU & memory):

dockai analyze <container-name> --perf 60

🧩 CLI Commands Overview

DockAI provides a comprehensive set of command-line commands to interact with the tool efficiently. Below is an overview of the main commands, their usage, and examples.

`analyze`

Analyze logs from a specified Docker container.

Usage:

dockai analyze <container-name> [options]

Options:

--since <duration>: Analyze logs since the specified time (e.g., 15m, 1h).
--tail <number>: Number of log lines to fetch from the end.
--perf <seconds>: Collect CPU and memory performance metrics for the given duration.

Example:

dockai analyze my_container --since 30m --tail 5000
dockai analyze my_container --perf 60

`list`

List all running Docker containers or available analysis plugins.

Usage:

dockai list

`config`

Manage DockAI configuration settings.

Subcommands:

set <KEY> <VALUE>: Set a configuration key.
get <KEY>: Retrieve the value of a configuration key.
list: List all current configuration settings.

Examples:

dockai config set DOCKAI_MODE cloud
dockai config get DOCKAI_MODE
dockai config list

`help`

Display help information about DockAI commands.

Usage:

dockai --help

Example:

dockai config --help

Quick Reference Table

Command	Description	Example
`dockai analyze`	Analyze container logs	`dockai analyze my_container --since 15m`
`dockai list`	List running Docker containers	`dockai list containers`
`dockai config set`	Set a configuration key	`dockai config set DOCKAI_MODE cloud`
`dockai config get`	Get a configuration key	`dockai config get DOCKAI_MODE`
`dockai config list`	List all configuration keys	`dockai config list`
`dockai --help`	Show help for commands	`dockai help analyze`

🔌 Plugin System

DockAI features a modular plugin architecture, allowing you to extend functionality without modifying the core system.

Plugins can hook into lifecycle events such as:

on_start — before analysis begins
on_logs_fetched — after logs are collected
on_ai_response — after AI model produces output
on_finish — when analysis completes

Example plugin structure:

# ~/.dockai/plugins/demo/plugin.py
class DemoPlugin:
    def on_finish(self, ctx, result):
        print("[demo] saved usage stats")

⚙️ Configuration

DockAI reads configuration from a simple JSON file:

~/.dockai_config.json

Example configuration:

{
  "DOCKAI_MODE": "local",
  "DOCKAI_OLLAMA_MODEL": "qwen2.5:7b-instruct"
}

You can update settings with:

dockai config set DOCKAI_MODE cloud
dockai config list

📊 Telemetry (Local Analytics)

DockAI includes a local telemetry system powered by SQLite — no data ever leaves your machine.

Each analysis records:

Container name
Duration and latency
CPU and memory statistics
AI model used

You can inspect telemetry data manually via:

sqlite3 ~/.dockai/usage.db
SELECT * FROM usage;

📚 Learn More

🌐 Website
🐍 PyPI
💻 GitHub

📘 Note: This documentation is a work in progress. More pages such as Getting Started, Telemetry, and Plugin Development will be added soon.