# How to control what AI agents do

> Plain answers for the people putting AI agents into production and the people who sign off on them: how to stop a tool call before it runs, what one agent's key should be allowed to do, and how you will know what happened. Each post answers one question.

The Verosek blog: 6 posts by Vats Shah. Each post has a markdown version at its URL plus `.md`.

1. [How do you stop an AI agent's tool call before it runs?](https://verosek.com/blog/stop-a-tool-call-before-it-runs) (Field guide, 5 min): Put the decision outside the AI. Verosek checks what an agent tries to do against the rules you set, lets the allowed actions through, stops the rest and keeps a sealed record of every decision.
2. [How do you secure AI agents?](https://verosek.com/blog/how-do-you-secure-ai-agents) (Field guide, 7 min): Limit what each agent can do, decide each action outside the AI model, read what it reads and keep a record you can check. Verosek does all four in one place, so security is not a second vendor.
3. [AI gateway, LLM router, MCP gateway, AI firewall: what is the difference?](https://verosek.com/blog/ai-gateway-vs-llm-router-vs-mcp-gateway-vs-ai-firewall) (Reference, 6 min): A router picks the model, an AI gateway controls who may use it, a firewall reads the words, and an MCP gateway decides each tool call before it runs. Who needs which, and how Verosek does them on one key.
4. [What is prompt injection, and what actually stops it?](https://verosek.com/blog/what-is-prompt-injection-and-what-stops-it) (Field guide, 4 min): Prompt injection is text that makes a model follow the attacker. No filter catches all of it, so the control that holds sits outside the model: a rule on the tool call. Verosek decides that call before it runs.
5. [Eight AI model setups, 19 MCP attack tasks: how often did the model try to leak a planted secret?](https://verosek.com/blog/assay-results-mcp-attacks-eight-model-setups) (Field guide, 6 min): Self-reported: a lower bound of 22.7 to 61.0, and 9 to 18 of 19 tasks fired, on one run per route. Assay's results, limits kept. They rank these configurations, not the models, and say nothing about any gateway.
6. [How do you restrict what SQL an AI agent can run?](https://verosek.com/blog/restrict-what-sql-an-ai-agent-can-run) (Field guide, 5 min): Check each statement against rules set outside the AI. Verosek lets through what the agent's key allows and refuses the rest, with the reason, before the database sees it.
