
This is a comprehensive OSINT framework that wraps 11 investigation tools (Holehe, Sherlock, Sublist3r, etc.) behind an AI agent that can chain them intelligently based on what it finds. You can run it three ways: an interactive REPL where you type natural language queries, direct CLI commands for scripting, or as a Python API. It handles emails, usernames, domains, IPs, and phone numbers with async execution and hard timeouts to prevent hanging. The AI piece is legitimately useful here since OSINT workflows naturally branch based on results, like finding an email in a breach then pivoting to username searches. Works with Claude or local Ollama models. The main friction is installing all the external dependencies and grabbing API keys for services like Shodan and HaveIBeenPwned to unlock full functionality.
npx -y skills add aradotso/security-skills --skill openosint-ai-osint-framework --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.