Products | DIMAGGI AI

What DIMAGGI AI builds: Tenwa (autonomous network intelligence), Caliber (tests and develops human reasoning and judgment), and Tool Guard (the runtime policy firewall for AI agents). Three separate products, not a platform.

Tool Guard

Tool Guard is DIMAGGI AI's runtime governance engine for AI agents. It evaluates every tool call, returns an ALLOW, DENY, or ESCALATE verdict, and records each decision in a tamper-evident SHA-256 hash-chained audit trail. Available in two editions: open-source Core and Enterprise.

Tool Guard Core (open-source, Apache 2.0)

The policy decision engine. Evaluates every tool call against your policy rules and returns a verdict before execution. Produces a tamper-evident, hash-chained audit trail verifiable with tg verify.

Tool Guard Enterprise

Adds the inline MCP enforcement gateway that intercepts tool calls and acts on the Core verdict before they execute. Includes role-based access control, PII redaction, signed Evidence Packs, and the management dashboard.

Shadow Mode

Deploy in shadow mode to observe all agent tool calls without blocking — understand your governance coverage before enforcing policies.

Enforcement Mode

Activate enforcement to block policy violations in real time. Every denied call produces a hash-chained audit record.

Policy Types

Explore Tool Guard

Tenwa

Catches faults in autonomous networks before they cascade - and fixes them itself, with every fix proven safe before it touches production. Tenwa watches live telemetry from the entire network fabric and rehearses every fix on a digital twin before it executes through the vendor's own API.

tenwa.ai

Caliber

Scores how well you actually think - and shows you exactly which parts of your reasoning are weak. Caliber grades written work across seven dimensions of judgment, from evidence grounding to defensibility, using a judge model rather than a fixed rubric.

caliber.dimaggi.ai