Built this after noticing that SOC analysts using AI tools were getting inconsistent results from single prompts.

The idea: instead of one vague prompt, break investigations into structured steps where each step builds on the previous output.

Now covers 8 workflows: phishing, PowerShell, ransomware, identity compromise, VPC logs, URL analysis, alert triage, and alert explanation.

Just added one-click "Run in ChatGPT" and "Run in Claude" buttons.

Free, no signup, no account needed.

Would love feedback from anyone working in security or building with LLMs.

https://www.socworkflows.com