SHELLOG

Your shell remembers. Now it understands.

Open-source shell history analytics with AI-powered pattern recognition. Discover hidden workflows, optimize command sequences, and audit security across your entire command-line history.

MIT License Bash / Zsh / Fish Rust Native AI-Powered
shellog analyze
$ shellog analyze --period 30d
Analyzing 4,827 commands from last 30 days...
 
Top Workflow Patterns:
  1. git pull -> npm install -> npm test
  2. docker build -> docker run -> docker logs
  3. cd project -> code . -> npm start
 
Optimization Suggestions:
  ! Create alias 'gpt' for git pull && npm test
  ! 23 redundant cd commands detected

Cognitive Shell Analytics

Transform your command history from a log file into actionable intelligence

Workflow Discovery

Automatically detect command sequences and recurring patterns. Identify your actual workflows, not just individual commands.

AI Pattern Recognition

Machine learning models trained on millions of shell sessions identify optimization opportunities and suggest aliases.

Security Auditing

Detect sensitive data in command history, find dangerous patterns, and identify potential credential exposure.

Productivity Metrics

Track command frequency, error rates, and time-based patterns. Understand your shell usage across projects.

Context Clustering

Group commands by project, directory, or semantic context. See how your work naturally organizes itself.

Cross-Shell Sync

Unified history across Bash, Zsh, and Fish. Your patterns follow you regardless of shell preference.

System Architecture

A modular design built for extensibility and performance

Shell Hooks

Bash / Zsh / Fish

SHELLOG Core

Rust Engine

SQLite

Local Storage

Pattern Engine

Sequence Mining

ML Analyzer

Classification

Security Scanner

Audit Module

Report Generator

Visualization

CLI Output

Terminal / JSON / HTML

How It Works

From raw history to actionable insights in four steps

Capture

Shell hooks capture every command with metadata: timestamp, directory, exit code, duration.

Index

Commands are normalized, tokenized, and stored in a local SQLite database with full-text search.

Analyze

Pattern recognition algorithms identify sequences, clusters, and anomalies in your usage.

Optimize

Receive actionable suggestions: aliases, scripts, security fixes, and workflow improvements.

50K+
GitHub Stars
2M+
Commands Analyzed
15ms
Avg. Capture Latency
100%
Local & Private

Get Started in Seconds

Choose your preferred installation method

Cargo (Rust)

Recommended for most users

cargo install shellog

Homebrew (macOS)

For macOS users

brew install einix/tap/shellog

APT (Debian/Ubuntu)

For Debian-based distros

apt install shellog

Ready to Understand Your Shell?

Join thousands of developers who have optimized their command-line workflow with SHELLOG.