TealTiger is the first open-source AI agent security SDK with client-side guardrails, policy enforcement, and audit logging. It wraps 7 LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, Mistral, Cohere) with consistent security — all running in your SDK with no servers required.Key capabilities:TealEngine: Deterministic policy enforcement with 3 rollout modes TealGuard: PII detection, prompt injection, content moderation in < 5ms

Wafaa.io — Turn scattered DMs into contracts you can control and prove.

Turn scattered DMs into contracts you can control and prove.

Hey everyone 👋 I'm the maker of TealTiger. We built TealTiger because we kept seeing the same problem: teams shipping AI agents to production with zero security controls. Not because they didn't care, but because every existing solution required deploying servers, proxies, or sidecars that nobody had time to set up. So we asked: what if the entire security layer lived inside the SDK itself? TealTiger gives you policy enforcement, PII detection, prompt injection blocking, circuit breakers, cost governance, and compliance-ready audit logging — all with npm install tealtiger or pip install tealtiger. No infrastructure. No config servers. Just import and go. We cover 7 out of 10 OWASP Top 10 for Agentic Applications through SDK alone, support 7 LLM providers with full feature parity in both TypeScript and Python, and every security check runs in under 5ms. It's fully open source (Apache 2.0). Would love your feedback — what security concerns do you face when deploying AI agents to production?

Wafaa.io — Turn scattered DMs into contracts you can control and prove.

Turn scattered DMs into contracts you can control and prove.

TealTiger

Open-source AI agent security SDK — zero infrastructure

Open-source AI agent security SDK — zero infrastructure