Authorization and scope come before testing.

1. Submitter authorization

You may submit only an application you own or have explicit permission to assess. The submitter must attest to that authority. SafeForProd may decline or stop work when ownership, authorization, or target identity is unclear.

2. Free passive scan

The free scan is limited to the submitted live URL, the public files that page references, response headers, and browser-visible configuration or page-load signals. It may identify public source maps, credential-like patterns, referenced endpoints, frameworks, and deployment metadata.

It does not guess hidden paths, request unreferenced admin or backup files, call discovered APIs, submit forms, create accounts, query data, use discovered credentials, introspect GraphQL, exercise authorization, verify RLS, or test rate limits.

3. Planned Authorized source scan

When launched, source intake and active runtime checks will require an account, written authorization, verified domain control, a recorded target list, and an approved scope. ZIP and read-only repository access will be limited to the selected project. Runtime validation will use read-only methods where possible, strict request and concurrency budgets, redacted evidence, and explicit stop conditions.

4. Always prohibited

• Denial-of-service, load, stress, or resource-exhaustion testing.
• Destructive payloads, persistence, malware, social engineering, or phishing.
• Accessing, modifying, retaining, or publishing customer data beyond the minimum evidence agreed in scope.
• Testing third-party systems, shared infrastructure, or unrelated tenants without their separate authorization.
• Expanding scope based only on a hostname, repository reference, or credential discovered during a scan.

5. Evidence and sensitive values

Reports minimize and redact tokens, personal data, and proprietary source. A public or publishable key is not labeled a secret leak without provider-aware evidence. Discovered values are not used during the free scan. Retention and deletion terms for source intake must be agreed before the planned Authorized source scan launches.

6. Stop conditions

Testing stops when the target behaves unexpectedly, availability may be affected, scope becomes ambiguous, sensitive third-party data appears, or the customer withdraws authorization. Work resumes only after the condition is understood and the scope remains valid.

7. Questions

Questions about a proposed scope should be sent to hello@safeforprod.com before submitting source or test credentials.