SpiderCob: Open-source AI firewall stopping data leaks before they ship

Hey PH! 👋 I'm Sumeet, the solo dev behind SpiderCob. I built this after watching a $2M data breach unfold at a company I consulted for — caused by a developer pasting an AWS key into a ChatGPT prompt. The security team had no visibility into what was going into the LLM. SpiderCob started as "what if I had a transparent proxy that scanned everything before it hit OpenAI?" and grew into a full DLP platform over the past 6 months. What makes it different from Nightfall / Symantec DLP: - Completely free and open-source (Apache 2.0) - Deploys in 5 minutes, not 6 months - AI-native — built specifically for the LLM prompt leakage problem - Fine-tuned ML models (94% accuracy) to kill false positives Live demo: https://spidercob.com/demo — paste any text or upload a file and see it scan in real time. No sign-up needed. Self-host: docker compose up -d — full stack runs on a $6/mo VPS. Would love feedback on: 1. What integrations matter most to you? (Slack, Jira, Confluence, MS Teams?) 2. What compliance framework would make this a no-brainer for your team? (SOC2, ISO 27001, HIPAA?) 3. Is the free self-hosted model right, or would you pay for a managed cloud version? Ask me anything!