AI Agent Lifecycle
AI agents act on their own — calling tools, reading untrusted content, moving data, and reaching external services. Netzilo AIDR governs what they do, checking every prompt, model response, and tool call against your policy and stopping prompt injection, credential exfiltration, capability hijacking, and data leakage before they complete.
How you connect Netzilo to an agent depends on where that agent is in its lifecycle and how much control you have over its code.
The lifecycle
| Stage | What it covers | How Netzilo governs it |
|---|---|---|
| Development | Agents your team writes and runs in its own process | The SDK, embedded directly in the agent |
| Pre-deployment & run-time | Agents already built — coding-agent CLIs, packaged tools, and any AI activity on a machine or in a browser | Plugins, the Netzilo client, and the Netzilo browser extension |
Agent types
Custom Agents — you build them
Agents written by your team. Because you own the code, you embed Netzilo directly in the process with the SDK (Python or TypeScript): one call at startup, then every framework tool call and LLM prompt/response is governed in-process. This is the development-stage integration.
Coding Agents — you run them
Third-party coding-agent CLIs you use but don't build — Claude Code, OpenAI Codex CLI, Google Gemini CLI, OpenClaw. Netzilo registers a plugin with each so their tool calls and prompts are checked against policy. This governs them at pre-deployment and run-time, without changing their code.
General-Purpose Agents — anything running on a managed machine
Any AI agent running on a desktop or server — Claude Code, Claude Cowork, OpenAI Codex, OpenClaw, and every other coding agent, desktop assistant, or in-house automation. Installing the Netzilo client monitors every agent on that machine across its full activity — file system, processes, MCP tool calls, LLM calls, and network — and the Netzilo browser extension brings the same AIDR functionality to browser-based agents. This is the broadest run-time integration: it covers agents you neither built nor have a dedicated plugin for.
These approaches compose. A team commonly embeds the SDK in the agents it builds, installs the client on the machines its agents run on, and registers plugins for the coding-agent CLIs its developers use — all reporting to the same Netzilo policy and dashboard.