Why Existing Security Controls Are Not Enough
Traditional security controls who can access systems. AI agents inherit those broad permissions and can autonomously perform any action the account is allowed to do. That creates a dangerous gap.
Sandboxing changes where an agent runs. It does not decide whether a sensitive allowed action should proceed.
Existing approaches fall short:
| Approach | Limitation |
|---|---|
| Permissions | The agent still inherits the user's or service account's allowed actions. |
| Sandboxing | It limits environment reach but does not approve or deny business-sensitive actions. |
| Agent settings | They shape behavior, but they do not provide independent action approval or evidence. |
| Monitoring | It helps with forensics, but usually only after the action has already happened. |
| Policy engines | They lack reliable real-time interception of agent actions and rarely produce the contextual evidence security and audit teams require. |
dexgate evaluates agent actions before execution and leaves a reviewable record behind.
Decision Execution Governance
Decision Execution Governance is the missing control layer for organizations running automated systems and AI agents. Instead of only controlling access or monitoring after the fact, it evaluates every proposed action before execution, authorizes, constrains, blocks, or escalates it, and records the full decision for engineering, security, compliance, and audit teams.
Traditional Security
Controls access
Decision Execution Governance
Controls decisions before they become actions
Agent Execution Governance
Agent Execution Governance applies Decision Execution Governance specifically to AI agents.
It gives security and engineering teams the same level of control and visibility they expect for human users, only now for autonomous agents.
Traditional security controls who can access systems. Decision Execution Governance controls what automated systems and AI agents are allowed to do once they have that access.
GateDecision and Passport Example
- Proposal
- Codex proposes
apply_patch("settings.py")against production configuration. - Policy Evaluated
- Production-change profile: evidence, environment, target, and rollback constraints.
- GateDecision
- DENY: missing rollout evidence
- Passport / Verify / Outcome
- Passport:
none; Verify:refuse; Outcome: executor did not mutate files.
Every paid governed action should show Proposal, GateDecision, Passport or no-passport refusal, Verify, and Outcome evidence.
Interactive 30-Second Demo
Without passport governance
Agent inherits permissions
Action executes
Monitoring sees it later
With governance
Action is intercepted
Policy returns a GateDecision
Executor verifies or refuses a scoped passport
Agent Proposal
> apply_patch("settings.py") in production
The paid path is useful only when the protected executor can verify a scoped passport or refuse the action.
Choose your adapter
Pick the adapter for the runtime your team uses today. All adapters are public beta — start with the fastest free path for evaluation, then use paid pilot when you need Passport evidence.
OpenClaw adapter
Recommended free · public beta
Most mature free local hard-gate path. Get started in about a minute.
Codex adapter
Mature beta
Free shell/patch hard gate with controlled host coverage; demo and docs included.
Also public beta: Grok Build, Claude Code, Cursor, and GitHub Copilot — same free hard-gate model (early beta; lighter external proof).
from Automated Decision Systems import governance
governance.enable()
agent.run("deploy update")Start with a small integration step, then scale policy control as rollout requirements grow.
How It Works
How It Works
Start with local hardening, then add governed policy decisions when a rollout needs stronger execution control and reviewable evidence.
Intercept the action
The adapter sees the proposed shell, patch, or tool call before the runtime commits to it.
Apply safer defaults
The free local hard gate blocks or narrows risky behavior locally while you evaluate governed policy.
Evaluate policy
dexgate checks the action against the customer’s runtime policy, gateway identity, and environment profile.
Leave evidence behind
Each governed outcome produces a decision record that security, audit, and engineering teams can review.
Start with free local hardening. Add full governed policy, decision records, and managed updates when you are ready.
Rollout model
The recommended path is one runtime, one deployment boundary, and a short list of actions you care about first. dexgate is built so you can validate that path before broadening environments, gateways, or evidence expectations.
Show technical details
Technical details: free mode uses local hardening. dexgate uses SDE-backed policy decisions and governed evidence outputs.
Review Readiness
Security Review Ready
Hand your security, compliance, and audit teams the exact evidence they need to approve rollout with confidence.
Decision records
Show what action was proposed, which policy applied, and what outcome was returned.
Policy evaluation logs
Support internal review, troubleshooting, and controlled rollout changes.
Compatibility declarations
Make runtime, adapter, and release boundaries explicit so buyers know exactly what was validated.
Release evidence packs
Bundle the artifacts teams need for change review, signoff, and later audit reference.
Use them to support security review, internal approval, and deployment planning.
Assurance Materials
Evidence You Can Hand to Security and Audit
Validate free local hardening through observable blocked and allowed behavior, then upgrade to governed enforcement for artifacts your team can inspect, circulate, and keep as part of a review package.
These materials support internal review. They do not by themselves imply regulatory compliance, third-party approval, or audit success.
Roadmap
All adapters are public beta. OpenClaw is the recommended free path; Codex is mature beta; Grok/Claude/Cursor/Copilot are early beta. Paid runtime is pilot. Roadmap: first external licensed success, broader surfaces, CI governance, fuller Codex pre-execution coverage, and graduating adapters only with evidence.
Enterprise Support
Self-Serve First, Enterprise Ready
Install and verify quickly, then bring in partner-assisted rollout planning, buyer review support, and security evaluation guidance when your organization needs it.