@shv_founder ↗
Services / API security audit Moscow · Worldwide

API security audit: tokens, IDOR, oversharing

REST/GraphQL: auth, scopes, rate limits, mass assignment, injection. Frontends change — the API stays the trust boundary.

Why it matters

Most breaches don’t start at the pretty UI — they start at an endpoint that skipped object ownership checks or returned one field too many. Without an audit you learn that from a customer, a regulator, or post-incident logs. You pay to close holes before production traffic, not to clean up after.

What we do

In this engagement:

  • Endpoint and role map — who should reach what, where the model diverges from reality
  • AuthZ and IDOR — ID swapping, foreign objects, vertical and horizontal privilege escalation
  • Tokens, sessions, scopes — lifetime, refresh flows, leaks in logs and clients
  • Rate limits and anti-abuse — brute force, enumeration, expensive ops with no brakes
  • Schemas, mass assignment, over-fetch — extra response fields, writing forbidden attributes, weak input validation

How we work

We start from OpenAPI/Postman, or capture traffic from stage/prod. Critical path first: auth → object-level rights → input/output → limits. Every confirmed finding comes with a PoC request and business/data risk — not CVSS theater. You get a report, fix priorities, and contract-level recommendations. Timeline is set after a short scope review.

FAQ

How long does an API security audit take and what do you need from us?

Typically we need a spec (OpenAPI/Postman), a test environment, and a role list. Timeline depends on endpoint count and auth depth — we lock it after a short scope review, not a month-long audit for its own sake.

Is this just a report, or do you help close the holes?

The core deliverable is a prioritized report with PoCs and contract-level recommendations. If you want, SHV can stay for the fix: patch authZ, tighten schemas, rate limits, and re-check critical paths. Audit and remediation can be scoped separately.

Next / Your project

Share OpenAPI/staging — we’ll probe endpoints.