Skip to main content
Mintive is the new name for MintMCP.Read the announcement

Agent security

AI agents have access that traditional applications never needed, and that access creates security challenges.

What agents can access​

Access typeExamplesRisk
Prompts and attachmentsContracts, HR spreadsheets, customer exports pasted or attached in claude.ai and CoworkSensitive data exposure
TerminalBash commands, package installation, system configArbitrary code execution
File systemSource code, .env files, SSH keys, configsCredential exposure
MCP toolsCRM, email, databases, APIs, cloud servicesData exfiltration, unintended writes
NetworkHTTP requests, database connectionsUnauthorized access

Why security tools can't see what AI agents do​

You can't govern what you can't see. Traditional security tools don't capture agent behavior because:

  • Agent actions span multiple systems and interfaces
  • Decision-making happens inside the AI model
  • Tool usage changes dynamically based on context
  • Most monitoring tools were built for human behavior

AI agent security risks​

Sensitive data in prompts​

People attach and paste the material they're working on, so prompts in claude.ai and Claude Cowork carry customer lists, contract terms, salary data, and financial forecasts. Claude inference hooks send each governed prompt to Agent Monitor, where rules can block it before it reaches the model.

Actions through connected tools​

An agent with MCP tools can send email, update CRM records, or export data on a user's behalf. The MCP gateway controls which tools each team gets, and Agent Monitor records every call.

Sensitive file access (coding agents)​

Agents can read files you'd rather they didn't:

  • Credentials: .env files, SSH keys, API tokens, database passwords
  • Production config: Connection strings, deployment scripts, IaC files
  • Business data: Proprietary code, customer data, financial information

Command execution (coding agents)​

With bash access, coding agents can:

  • Read secrets from files or run kubectl get secrets
  • Deploy to production without approval workflows
  • Install packages that introduce vulnerabilities
  • Call cloud APIs with available credentials

Prototyping vs production​

The security posture that works for prototyping doesn't work for production:

AspectPrototypingProduction
CredentialsHardcoded values acceptableSecret management, rotation
AccessBroad permissions for explorationLeast privilege, role-based
Change managementDirect deploymentApproval workflows, testing
MonitoringMinimal loggingComprehensive audit trails
Agent accessFull system for experimentationRestricted to specific tasks

The challenge is letting people experiment with agents while keeping production systems and sensitive data on approved paths.