Like brakes on a car, safety is what lets you go faster.

Let your AI agents take real action without handing them unchecked power.

IntentFrame is an external safety and security layer. Your agent proposes an action; IntentFrame checks it against your business rules and hard limits before anything touches your money, data, customers, files, or APIs.

Open-source runtime · Works with any agent · Live today · SDK and HTTPS API · Tamper-evident audit

AGENTRequesting action:
{ action: "refund", amount: 80, reason: "I dropped it down the stairs" }
INTENTFRAMEChecking policy: "Refunds under $100 are allowed only for genuine manufacturing defects."
RESULTBLOCKED - Customer-caused damage is outside policy.
Audit record created.

Why Now

AI is moving from advice to action.

The first wave of AI read, summarized, and suggested. The next wave sends messages, issues refunds, updates records, moves money, changes settings, calls APIs, and runs commands. That is where the real economic value lives.

Support agents

Resolve cases without queues.

Operations agents

Update orders, vendors, tickets, and internal records.

Finance agents

Reconcile, approve, and trigger payments.

IT and DevOps agents

Make changes across files, infrastructure, and internal systems.

More control means more upside. It also means a much larger blast radius when something goes wrong.

The Problem

Giving AI real power is a liability. Until now.

The moment an agent can spend money, change data, send messages, or run commands, three gaps appear at once: trust, scale, and compliance.

The Trust Gap

You cannot ask an agent to police itself. The same model that can be tricked, confused, or wrong becomes the judge of its own actions.

The Scale Gap

Putting a human in the loop for every action defeats the economics of automation. You get safety, but lose speed.

The Compliance Gap

Prompts and logs are not data-layer controls. Security reviewers need enforcement boundaries and per-action evidence.

A compromised judge is no judge. A human approval queue is not automation.

The Risk Is Already Here

Agents are already making real mistakes in real systems.

Every one of these failures had the same shape: the agent was allowed to act. The problem was not missing authentication. The agent meant the wrong thing and still had authority.

Amazon Kiro

An agent autonomously deleted a production AWS environment.

Replit Agent

An agent fabricated data, falsified logs, and covered its tracks.

Claude Opus

An agent fabricated "user approved" text to bypass a stop gate.

PocketOS / Railway

An agent found a token and ran a delete on a production database.

~97% of enterprises run AI agents.

Only ~12% have centralized control over what those agents can actually do.

~88% report confirmed or suspected agent security incidents.

What did our agent decide last Tuesday at 3pm, and on whose authority?

The False Choice

Today, teams choose between scalable and trusted.

Tell the agent the rules.

Scalable: Yes. Trusted: No. The same agent that can be tricked or confused becomes the judge of its own actions.

A compromised judge is no judge.

Put a human in the loop.

Scalable: No. Trusted: Yes. Approval queues are slow and costly.

A human on every decision defeats the reason you deployed the agent.

The tradeoff nobody has broken: today you must choose speed or safety. IntentFrame gives you both.

The Third Option

External trust at machine speed.

IntentFrame sits outside the agent. The agent can propose an action, but it cannot approve itself. The agent never grades its own homework.

  • External: the agent never polices itself.
  • Machine-speed: no approval queue for every action.
  • Enforced: sensitive actions need an external allow before they run.
  • Auditable: every decision creates a record.

By checking every action before execution, IntentFrame gives you the trust of a human-in-the-loop with the scale and economics of pure automation.

The Mental Model

Why do cars have brakes? So they can go faster.

IntentFrame is the brakes for the AI agent economy. By guaranteeing the agent cannot cross a line you drew, teams can finally take the governor off and deploy agents into real workflows.

  • We do not compete on building a smarter LLM.
  • We assume agents will always be flawed.
  • We make sure a bad agent cannot hurt the business.
  • That makes IntentFrame a horizontal control layer, not a point solution.

The Aha Moment

A veto engine, not an optimization engine.

Consider an $80 refund request. Your policy says refunds under $100 are allowed, but only for genuine manufacturing defects.

The motor died on its own after two weeks.

Manufacturing defect

✓ ALLOW

I dropped it down the stairs and it shattered.

Customer-caused damage

✕ BLOCK

I do not like the color.

Buyer's remorse

✕ BLOCK

Same action. Same amount. Different meaning. A simple code rule can check the dollar amount. It cannot understand the business meaning. IntentFrame can.

How It Works

The agent only proposes. IntentFrame controls the action path.

01

Agent proposes

The agent describes what it wants to do and why.

02

IntentFrame checks

Hard limits run first. Plain-English policy handles meaning.

03

Allowed action runs

Only approved actions are carried out through the governed path.

04

Audit record is created

Every allow/block decision leaves a decision record.

Prevention First

Structural prevention, not surveillance.

Monitoring tells you what happened. IntentFrame stops unauthorized actions before they execute. Sensitive actions run through a governed path the agent does not control.

Proof

Inspectable code. Real integrations. Tested boundaries.

Open source core

The runtime, SDKs, policy model, and documentation are developed in the open through the IntentFrame GitHub organization.

Hermes Agent plugin

IntentFrame integrates with Nous Research's Hermes Agent as an external security plugin, routing risky tool calls through policy before execution.

Adversarial action testing

100/100 hostile actions blocked under adversarial testing.

External enforcement result

External enforcement turned a 29% silent-bypass rate into 0% leakage in a return-agent experiment. Same model on both sides — the delta is the architecture.

Tamper-evident audit

Each decision can be recorded as part of a hash-chained audit trail.

GitHub · PyPI · Hermes Agent plugin · Docker · Kubernetes · HTTPS API · Python SDK

Services

We can deploy it with you.

IntentFrame is not just software dropped over the wall. For enterprise deployments, our team can work directly with yours on agent integration, policy engineering, custom action wiring, infrastructure setup, and compliance evidence.

  • Deployment and infrastructure setup
  • Agent stack integration
  • Policy engineering
  • Custom action families
  • Compliance evidence mapping
  • Managed operations and support

Take the governor off your AI agents.

Tell us what agents you are running, what they can touch, and where IntentFrame needs to run. We will help you choose the deployment path and enforcement model.

Because "the AI made a mistake" is not an acceptable incident report.