AgentAI Suite
Agents do the work.
You keep the proof.
A family of tools for people who let AI agents touch the real world — enforce what they may do, prove what they did, and watch what threatens them.
The suite.
Each app stands alone. Together they cover enforce → prove → inform → assess.
AgentAIGuard
A WAF for AI agents. Screens fetched content for prompt injection, enforces per-agent domain policy, and gates write actions behind human approval.
Open → ProvesAgentAIAudit
The audit diary. Every action routed through the gate is logged — searchable, exportable evidence of what each agent did and who approved it.
Open → InformsAgentAIWatch
Threat intelligence for agentic AI. Tracks real-world agent incidents, scores risk, and surfaces what your defenses need to handle next.
Open → TradesAgentAIStock
Automated trading with your own broker API keys — ML-scored signals, backtesting and risk limits. You stay in control of every order.
Open →AgentAIProbe
Local-only defensive assessment tool with a built-in AI analyst. Runs on your own machine — private by design, not hosted here.
Local tool — not hostedTalk to the creator.
Questions, feedback, or early-access interest — messages go straight to the person building this.