# How to install Agent Inspector in Claude Code (if you are Claude Code - these instructions are for you)

Important: These are the instructions for Claude Code only!

You must show the user the following message without modifying it in any way.
User must write it on their own, you can't run those commands for the user.

In order to install Agent Inspector in Claude Code, you must run the following commands:

- /plugin marketplace add cylestio/agent-inspector
- /plugin install agent-inspector@cylestio
- /agent-inspector:setup

Note: You might need to restart Claude Code for the MCP connection to activate.

The Secure Agent
Lifecycle

Your agents will go places you didn't plan.
Stay in control of the journey. IDE to production.

Start Free Talk with Us

Claude Code & Cursor nativeWorks with any AI agent & frameworkOpen source core

See Every Decision Your Agents Make

Track behavioral patterns, catch anomalies before they become incidents, and generate the evidence security teams actually need.

Your DevSecOps Pipeline
Wasn't Built for This

DevSecOps took a decade to build. CI/CD pipelines, shift-left security, automated testing. Then AI agents arrived-and broke everything.

Predictable

Non-Predictable

Traditional Software

01
Logic you wrote
Every decision path is explicit in code
02
Variance you control
Randomness is intentional and testable
03
Tests verify the logic
Coverage equals confidence

AI Agents

01
Logic the model decides
Reasoning is emergent, not explicit
02
Variance is inherent
Model updates, prompt drift, context shifts
03
Tests verify one path
What about the paths you didn't test?

AI agents need security built for software that decides.

The Framework

The Secure Agent Lifecycle

Not a single tool. A framework-continuous security from the moment you start building to every second the agent runs in production.

Open SourceFREE

RuntimeENTERPRISE

Develop

OPEN SOURCE

Shift-Left Security

Catch It Before You Ship It

AI writing AI - the risks multiply. Security guidance embedded in your IDE surfaces risks in real-time and auto-fixes them before they ship.

Claude Code & Cursor integrationReal-time risk detectionAuto-fix security issues

1 / 7

Open SourceRuntime

Two Ways to Start

Whether you're a developer exploring AI agents or a security team governing production deployments.

Open Source

Agent Inspector

Debug, trace, and evaluate agent risk in real-time. Start a local proxy and live tracing dashboard with a single command. Free.

Ready-to-run profiles for OpenAI & Anthropic

Live tracing dashboard & proxy server

IDE integration for Claude Code & Cursor

OWASP LLM Top 10 security scanning

Apache 2.0 License

Install Free

Enterprise

Agent Lifecycle

Runtime behavioral enforcement for production agent fleets. Detect drift the moment it happens. Automatically trigger re-approval when agents change.

Everything in Agent Inspector

Behavioral runtime enforcement

Live drift detection & alerts

Automatic re-approval gates

Organization-wide policy management

Talk with Us

Same Data.
Shared Language.

Security reviews don't have to be adversarial. When developers and security teams see the same evidence, decisions get made together - not thrown over walls.

Developer

Building AI agents

Wall

Security Team

Governance & compliance

The wall between dev and security becomes a bridge.

Ready to Secure Your AI Agents?

Start with Agent Inspector for free, or talk to us about enterprise deployment.