Guardian Agent 2.0: Security Ops Meet Diagram and GRC

•7 min read•Product

Security findings are more useful when you can see what they affect. That idea sits at the centre of Guardian Agent 2.0, the open-source local security workspace that merges Guardian security operations with Context Cypher diagram and GRC workflows in one application.

The current published line starts at the 2.0 security workspace merge (2.0.0-alpha.2 through 2.0.0-alpha.4). You clone the repository, build locally, and keep control of the workspace on your machine.

What changed in the merge

Guardian Agent now brings workstation and environment visibility together with the visual threat modelling and GRC workspace that Context Cypher teams already know. Instead of jumping between separate tools for posture checks, diagrams and governance records, you work in one browser workspace that binds to loopback by default.

Context Cypher remains available as a standalone open-source tool. You can still install it on its own, and you can import existing Context Cypher files into Guardian Agent when you want the combined workspace.

What you can do today

Capabilities come from the current Guardian Agent README and release notes, not a wish list:

  • Check your workstation: review host posture, Microsoft Defender status and registered antivirus products, then propose and approve supported Defender scans.
  • Understand your environment: review passive local network observations or enrolled AWS EC2 and security-group inventory, preview a snapshot, and turn it into an editable system diagram.
  • Model your systems: drag-and-drop node toolbox, security zones, embedded node and connection data, automatic zone colours, layout tools, built-in examples and Context Cypher imports.
  • Analyse threats with AI: ask security questions, analyse selected architecture and generate proposed diagrams with your chosen provider. Review results before you apply them. Options include OpenAI, Anthropic, Gemini, xAI and Ollama.
  • Connect diagrams to GRC: sync diagram nodes into the asset register, link risks to diagram elements, and manage controls, assessments, evidence and statements of applicability.
  • Review and report: investigate findings, record review decisions, inspect activity, and export diagrams, portable project files and GRC reports.
  • Work with AI assistants: give Codex, Claude Code and other compatible clients scoped access through MCP, CLI and HTTP interfaces, with approvals kept in the loop.

Diagram plus GRC in one place

Context Cypher's GRC workspace sits beside your diagrams inside Guardian Agent. That means assets, findings and risks can stay connected to the visual model instead of living in a separate spreadsheet or ticket pile.

  • Assets, findings and risks with scoring and treatment tracking
  • Compliance, controls, assessments, evidence and statements of applicability
  • Governance and risk management plans
  • Threat profiles, third parties, initiatives and incidents
  • Dashboards, reporting and workflow configuration

Open the shield icon in the Systems editor to switch to GRC. Use Sync from Diagram when you want diagram elements in the asset register.

Two ways to work

Use Guardian on its own

The browser workspace covers Protection, Environments, Findings, Systems, Activity, Integrations and Settings. Core monitoring, manual diagrams and GRC work do not require an AI account. For built-in AI analysis, configure a provider and model in settings. Provider keys stay in backend memory and are re-entered after a backend restart.

Connect your preferred assistant

Enroll a separate assistant credential in Settings, choose its permissions and project access, then point Codex, Claude Code or another compatible client at the MCP, CLI or HTTP interface. Credentials can expire or be revoked, and supported response actions still follow Guardian's approval process.

Get started locally

Requires Node.js 24.14 or later. From the project directory:

npm ci npm run build npm run init npm start

Open http://127.0.0.1:3000. Guardian binds to loopback on your machine. A practical first pass is: run a Protection check, open Systems with a built-in example, switch to GRC, then try Environments collect, preview and create editable system when you are ready to model your own environment.

Where to read more

Important: Guardian Agent is open source under Apache 2.0. AI-assisted analysis is designed to augment human expertise, not replace it. Validate findings before you act on them. Review our Terms of Service.

Back to Blog