Zero Trust, from inside the workflow
Network tools watch agents from the outside. Methos runs the agents, so it can check every step, tool and data request from the inside, where it knows exactly what the agent is trying to do.
SECURITY TODAY
The basics, done properly
- Single sign-on and managed authentication
- Role-based access for owners, admins and members
- Each organization's data kept separate from every other
- Approved model list enforced per organization
- Full session and tool-call history
THE ZERO TRUST MODULE
Never trust an agent by default. Verify every action.
The Zero Trust module treats each agent as its own identity with the least access it needs. Every request is checked against policy at the moment it happens, not once at login.
An example bulk export is blocked for excess access; a single-customer lookup passes all five security gates.
Example
Identity for every agent.
Each agent and sub-agent gets its own short-lived identity, tied to the person it acts for.
Policy at every step.
Each tool call and data request is checked against policy before it runs.
Least access by default.
Agents get only the tools and data their task needs, and lose them when the task ends.
Data protection.
Sensitive data is detected and kept out of places it shouldn't go.
Continuous monitoring.
Unusual behavior is flagged and can stop a run.
Human approval where required
Risky actions wait for a person to sign off before they run.
FRAMEWORK MAPPING
Mapped to the frameworks your reviewers use
The module is being designed against federal Zero Trust guidance, including NIST SP 800-207 and the CISA Zero Trust Maturity Model, so your security team can map its controls to the questionnaires they already answer.
YOUR SECURITY STACK
A complement to your existing tools, not a replacement
Keep your network and endpoint security. Methos adds control and evidence for the agents that run on Methos.
Compliance roadmap
Methos is built for regulated environments. We're pursuing independent, third-party attestations so your security and risk teams can verify our controls, not just take our word for it.
SOC 2 Type II
Independent CPA audit of our security controls over time.
In progressISO/IEC 27001
Internationally recognized information security management system.
In progressISO/IEC 42001
The international standard for AI management systems, covering how AI is governed, monitored, and overseen.
In progressWhat's in place today
- Encryption in transit and at rest
- Role-based access control with SSO and MFA
- Logs of agent actions, tool calls, and session history
- Guardrails and human-in-the-loop controls on workflows