$870M+
Your users are the target — malicious-signature phishing has taken more than $870M from over 760,000 wallets since 2023, inside flows like yours.
Put SafeClick inside your flow.
Why partner with us.
You get a verdict, enforced — only an allowed transfer reaches the broadcast path. No risk model to build, no score to interpret.
One layer, built for what’s next — the same decision layer for human wallets today and autonomous systems next.
Building one of these?
Non-custodial wallets
That want a judgment call on each transfer — not just spend caps.
Crypto payment teams
That want a risk check in the path before a transfer executes.
Embedded wallets
That own a signing flow and would rather integrate a verdict than build risk infrastructure in-house.
90-Day Enterprise Evaluation
See what SafeClick finds in your real transaction flow.
Run SafeClick against one agreed live workflow for 90 days. Before it starts, we agree expected decision volume and deployment scope and quote the evaluation against them. SafeClick returns evidence-backed Allow, Hold, or Block decisions while your existing controls remain authoritative.
Use the evaluation to measure coverage, review flagged activity, and decide where a SafeClick decision belongs in your production path.
Evaluate the decision layer against traffic you already understand.
Synthetic examples can show that an API responds. Your own workflow shows whether the verdict is useful.
During the evaluation, SafeClick observes an agreed transaction stream and produces a decision record with the evidence available at that moment. Your team can compare the result with its existing controls, investigations, and known outcomes.
Measure coverage
See how many transactions SafeClick evaluated, which workflows were represented, and where evidence was complete or unavailable.
Review risk
Separate activity that was flagged, investigated, customer-confirmed, or still unresolved. Each category keeps its evidence source and denominator.
Define the production path
Use the end-of-evaluation review to decide whether SafeClick should remain observational or move into a production integration, with terms set from the usage validated during the evaluation.
One company. One workflow. Ninety days of evidence.
- One approved company and one agreed transaction workflow
- An evaluation quoted against your expected decision volume and deployment scope
- An approved KYT configuration for the standard verdict pipeline
- Standard API onboarding and integration guidance
- Shadow decisions that do not alter the customer's execution path
- An end-of-evaluation report with coverage and outcome categories
- A technical and business review with the DetentLabs team
- Evaluation pricing expires after 90 days. Continuing into production requires a new commercial agreement based on validated usage and deployment requirements.
From application to evidence in four steps.
1
Apply with a real workflow
Tell us what the transaction flow does, which controls it uses today, its expected decision volume and deployment scope, and the failure you most want to catch.
2
Confirm the scope and quote
We agree on the workflow, data fields, expected decision volume, deployment scope, activation plan, and KYT provider configuration, then quote the evaluation against them. Applicants may use an approved supported provider account or request SafeClick-provided coverage.
3
Activate shadow decisions
The 90-day period begins with the first completed authenticated shadow decision. The application lapses if activation does not occur within 30 days of credential issuance. SafeClick evaluates the approved flow while the customer's existing controls remain authoritative.
4
Review the results
At the end of the evaluation, both teams review coverage, flagged activity, confirmed outcomes, unresolved cases, provider availability, and the production terms the validated usage supports.
KYT is part of the standard verdict.
Every Enterprise Evaluation runs with an approved KYT configuration. Applicants may bring an existing supported provider account or request SafeClick-provided coverage.
Provider eligibility, credential handling, and expected usage are confirmed during application review and reflected in the evaluation quote.
Production traffic, observational decisions.
SafeClick can evaluate data from an agreed live workflow while the customer's existing transaction controls continue to decide what executes. Shadow decisions are produced for measurement and review.
Shadow decisions are measurements for your review. They are not a guarantee that any threat will be detected or any loss prevented, and they do not replace your existing controls.
When a customer is ready for its production execution path to depend on a SafeClick decision, that work moves into a Production engagement with an explicit integration and responsibility boundary.
Send the transaction context required for a verdict.
Evaluation inputs are limited to the approved transaction, address, chain, and supporting metadata needed for analysis.
Never send private keys, seed phrases, wallet sessions, cookies, or unrelated customer data. SafeClick is non-custodial and does not require key material to produce a verdict.
Raw Evaluation data is retained for the agreed review and dispute window, then deleted as described in your written evaluation agreement and the SafeClick Privacy Policy. Raw data is used for model training only with written opt-in. Confirmed indicators of malicious activity may be retained to improve detection across the SafeClick network, subject to your written evaluation agreement, the SafeClick Privacy Policy, and applicable law.
An evaluation report with denominators, not a threat-count headline.
The report records the traffic SafeClick was asked to evaluate and distinguishes:
- completed shadow decisions
- flagged activity
- investigated cases
- customer-confirmed malicious activity
- unresolved activity
- provider-unavailable or incomplete-evidence events
The report records the Customer-provided workflow total, submitted volume, completed volume, sampling method, and coverage percentage so each result can be interpreted against the workflow that produced it.
From evaluation to production.
The Evaluation ends after 90 days. The Evaluation key is then disabled and evaluation pricing expires with the term. Continuing into production requires a new commercial agreement based on validated usage and deployment requirements.
Production terms are set from the usage observed during the evaluation: validated decision volume, deployment requirements, and enterprise scope. Customer-specific implementation, production enforcement, service commitments, and dedicated support are scoped in that agreement.
Enterprise Evaluation
Ninety days on one agreed live workflow, quoted against your expected decision volume and deployment scope. Your existing controls remain authoritative.
Production
Converted on validated usage: a production commitment plus metered decision usage, with terms set from observed volume, deployment requirements, and enterprise scope, and the required operational and service commitments.
Enterprise
Negotiated volume across multiple workflows, with custom policy, retention, support, and service terms for a larger deployment.
Evaluation FAQ
When does the 90-day period begin?
The Activation Date is the timestamp of the first completed authenticated shadow decision, unless both parties agree to another date in writing. The application lapses if activation does not occur within 30 days of credential issuance.
How is the Evaluation priced, and what counts as usage?
The Enterprise Evaluation is paid. Before it starts, you state the expected decision volume and deployment scope, and DetentLabs quotes the evaluation against them. Each completed ALLOW, HOLD, or BLOCK shadow decision counts as usage, including a completed decision with provider-unavailable or incomplete evidence. Authentication failures, malformed requests, internal retries, idempotent duplicates, and technical failures without a completed decision do not count. Evidence gaps are reported separately and handled during the end-of-evaluation review.
Does the Evaluation change how our transactions execute?
Your existing controls remain authoritative throughout the Evaluation. SafeClick returns observational decisions for measurement and review.
Can we use our existing KYT provider?
Applicants may bring an approved supported provider account. Provider eligibility, delegated use, credential handling, chain coverage, and available capacity are confirmed during application review.
What if we need SafeClick-provided KYT coverage?
Request it in the application. DetentLabs will confirm availability and reflect the approved provider configuration in the evaluation quote.
What happens if a provider is temporarily unavailable?
SafeClick records the provider state and preserves fail-closed decision semantics. A completed decision still counts as usage. Provider-unavailable and incomplete-evidence events are reported separately and handled during the end-of-evaluation review.
What happens after 90 days?
Evaluation pricing expires and the Evaluation key is disabled. Continuing into production requires a new commercial agreement: production terms are set from the validated usage observed during the evaluation, your deployment requirements, and enterprise scope, with decision usage metered from there. Larger deployments move to Enterprise terms.
Can SafeClick enforce decisions in production?
Pre-broadcast enforcement is available through a Production engagement, where the execution path, support, and responsibility boundary are agreed explicitly.
We are building an AI-agent payment flow. Can we apply?
Tell us about the flow. Agent-payment and x402 integrations are reviewed separately based on the current product surface and the party that controls execution.
Bring us one transaction flow worth evaluating.
Tell us how value moves, what controls it today, and which failure would matter most. We will review the workflow, KYT configuration, data boundary, and expected decision volume with you, then quote the evaluation.
Bring us a real payment flow.
Prefer email? partner@detentlabs.ai