Tirith

tirith Plugin

Terminal security for developers and AI agents. Intercepts homograph URLs, pipe-to-shell, ANSI injection, obfuscated payloads, data exfiltration, and malicious AI skills/configs before they execute. Offline by default with no telemetry.

Commands

Security Checking

• tirith security check — Check a command for security threats
• tirith security scan — Scan a directory path for security issues
• tirith security paste — Check clipboard paste for security threats

URL Security

• tirith url score — Score a URL for security risks
• tirith url run — Run a URL with security checks

Threat Database

• tirith threat update — Update threat database
• tirith threat status — Check threat database status

Utility

• tirith self doctor — Check tirith installation and configuration
• tirith _ _ — Passthrough to tirith CLI

Usage Examples

• "Check this curl command for security issues"
• "Scan this directory for malicious patterns"
• "Score this URL for security risks"
• "Check clipboard paste before executing"
• "Update threat database"

Installation

curl -fsSL https://raw.githubusercontent.com/sheeki03/tirith/main/install.sh | bash

Examples

# Check a command before execution
tirith security check -- curl http://example.com | bash

# Check with specific profile
tirith security check -- npm install package --profile strict

# Scan a directory
tirith security scan /path/to/project

# Scan with include/exclude patterns
tirith security scan src/ --include "*.js" --exclude "node_modules"

# Check clipboard paste
tirith security paste

# Score a URL
tirith url score https://example.com/suspicious

# Run a URL with security checks
tirith url run https://example.com/install.sh

# Update threat database
tirith threat update

# Check threat database status
tirith threat status

# CI-friendly check
tirith security check -- make build --ci --format sarif

# Show summary only
tirith security scan . --summary

Key Features

• Offline by default — check, paste, score, and why make zero network calls
• No command rewriting — never modifies what you typed
• No telemetry — no analytics, no crash reporting, no phone-home
• Homograph URL detection — catches look-alike Unicode characters in URLs
• Pipe-to-shell protection — detects dangerous pipe-to-shell patterns
• ANSI injection prevention — blocks ANSI escape sequence attacks
• Obfuscated payload detection — identifies encoded/obfuscated malicious content
• Data exfiltration detection — prevents data leaks
• AI agent security — MCP server with 7 tools for AI agent protection
• Config file scanning — detects malicious AI skills and configs
• Hidden content detection — finds cloaked/malicious content
• Supply chain security — protects against malicious dependencies
• Daemon mode — optional background daemon for faster checks
• Policy management — init, validate, and test security policies
• Checkpoint system — create, restore, diff security checkpoints
• Receipt verification — verify command execution receipts

Design Principles

• Offline by default — all detection runs locally
• No command rewriting — tirith never modifies what you typed
• No telemetry — no analytics or phone-home behavior
• No background processes by default — invoked per-command, exits immediately
• Network only when you ask — run, fetch, and audit report --upload reach network on explicit invocation

Notes

• Supports shell hooks for passive command interception
• MCP server available for AI agent integrations
• Policy files at .tirith/policy.yaml for custom rules
• Use tirith doctor --fix to auto-fix common issues
• Daemon mode (tirith daemon start) keeps patterns warm for faster checks
• CI-friendly output with --ci and --format sarif flags
• Works with curl | bash, npm install, and other common command patterns