Security testing services: what is in scope
Our application security testing services cover your application and its APIs, in a test environment, as written in the scope. The scope also says what is out of scope, and says it as plainly. Penetration testing, red-team exercises and social engineering are not part of this service, and nor are tests of physical or network infrastructure. If your customer or regulator names a standard, tell us. The scope says what the testing covers and what it does not.
The service is automated security testing: checks that keep running in your CI next to your other tests. A weakness that comes back shows up on the next change, not at the next audit. When you compare a security testing company with others, ask what keeps running after the report is delivered.
Kinds of checks: static, dynamic and access-rule tests
- Static: analysis of your code and its dependencies without running the application.
- Dynamic: checks against a running application in a test environment.
- Access-rule tests: tests that try to reach data or actions a role must not reach, for example one customer’s records from another customer’s account.
The scope chooses which of these run. Scenarios we start from are simple ones: sign-ins with invalid credentials are refused, session and cookie handling behaves as specified, and the data a response exposes is the data it should.
API security testing
For APIs the checks cover authentication, authorization per endpoint, input handling and data exposure in responses. Functional tests of your API rules, such as a user without a role getting a refusal, are written as ordinary tests; see API testing. Security testing looks for weaknesses nobody wrote a rule for. For web front ends, see web test automation.
Automated security testing in your CI
This is DevSecOps testing in its plainest form. Checks run on pull requests or nightly, as agreed. New findings block the build or warn, as agreed. Known findings are tracked, not reported again every run, so the signal stays readable.
How findings are verified and reported
Automated tools report false positives. An engineer checks each finding before it reaches the report. Each finding has a severity, steps to reproduce and the evidence. After a fix, the finding is retested.
How we handle access and data
Testing happens in test environments, with test accounts. Credentials are issued by you. Nothing is pointed at production without written agreement. Only tools you approve are used, and nothing leaves your environment if your rules say so. For the tools survey, see security testing tools.
What AI does and does not do in security testing
AI assistance is limited to drafting and repairing test code, and an engineer reviews each change before it is merged. People decide severity and what is reported. Testing AI features themselves is a different piece of work; see security testing of AI chatbots.
When you need something other than this service
If you need a penetration test, a certification or a compliance audit, this is not it, and the scope week would tell you so. Automated checks also need a test environment to run against and someone on your side to receive findings and fix them; without either, set those up first. If you are not sure what to check first, a QA audit looks at tests, pipeline and release process, and it says so when something else is needed. For behaviour under load, see performance testing.
From scope week to checks in your CI
- Security test scope: about one week. The written scope, rules of engagement and a first verified findings list.
- Pilot: two to six weeks. Checks in your CI, access-rule tests and a findings report.
- Ongoing: checks keep running and fixes are retested, through QA outsourcing, a dedicated QA team or engineers who join your team.
After each step you decide whether there is a next one. Prices are not published; send what the application does and the quote comes within one business day. Other test types are listed under test automation services.