AI Test Coverage Analyzer

AI Test Optimization

Provide the requirements and the current tests; the AI builds a coverage map, flags gaps by risk and proposes the missing tests.

Privacy: Requires a free Mutqan account. Your input is sent to the Mutqan AI service to generate the result; prompts and results are not stored — only your account's usage counts for the daily allowance. How the AI tools work

What is test coverage?

Test coverage measures how much of something — requirements, risks, code branches, input partitions — the tests exercise. Requirements coverage answers "is every acceptance criterion tested?", while code coverage answers "which lines ran?"; a high percentage of the second says nothing about whether the assertions were meaningful, so coverage is a map of gaps, not a score to maximise.

How to use AI Test Coverage Analyzer

  1. Paste the requirements or user story in the first box and the existing test cases in the second.
  2. Pick the output language and press Analyze.
  3. Review the traceability matrix: each requirement is marked full, partial or none, with the tests that cover it and what remains untested.
  4. Add the missing tests to your suite (the AI Test Case Generator can write them) and retire the redundant ones; download the report as CSV, Markdown or JSON.
Full counts 1, partial 0.5 and none 0, so the example scores (1 + 1 + 0) ÷ 3 = 67%, and the untested R-3 gets the missing test MT-1.
RuleTC-001delete succeedsTC-002referenced value errorCoverage
R-1Admin deletes an unreferenced value✓✓Fullcounts 1
R-2Referenced values are blocked✓✓Fullcounts 1
R-3Only Admins may deletenot covered by this testnot covered by this testNonecounts 0 · gap → MT-1
  • Full = 1: positive path and main negative or boundary cases tested
  • Partial = 0.5: only the happy path is tested (none in this example)
  • None = 0: no test, so a missing test is proposed
  • MT-1Non-admin DELETE returns 403 (for R-3; P0, security)

The tests that cover each rule in the example, one column per test. A test written for a neighbouring rule counts when it exercises that condition: refusing to delete a referenced value (TC-002) is R-1’s negative case.

AI Test Coverage Analyzer features

  • Requirement-to-test traceability matrix with full/partial/none coverage
  • Overall coverage percentage and risk per requirement
  • Prioritised list of missing tests
  • Flags redundant tests that map to no requirement

AI Test Coverage Analyzer example

Coverage

Input:

R: Admin deletes unreferenced value; referenced values blocked; only Admins.
T: TC-001 delete succeeds; TC-002 referenced value error

Output:

R-1 full (TC-001 + TC-002) · R-2 full (TC-002 + TC-001) · R-3 none
Coverage (1 + 1 + 0) ÷ 3 = 67% → MT-1 Non-admin DELETE returns 403 (R-3, P0, security)

Frequently asked questions about AI Test Coverage Analyzer

How is the coverage percentage computed?

Full coverage counts 1, partial counts 0.5 and none counts 0; the sum is divided by the number of requirements.

Why is a tested requirement marked partial?

Because full means both the positive path and its main negative or boundary conditions are tested. A requirement with only a happy-path test stays partial until its error and limit cases are covered too. A test written for a neighbouring rule counts when it exercises that condition: refusing to delete a referenced value is the negative case of deleting an unreferenced one.

Which tests are flagged as redundant?

Tests that map to no requirement, or that duplicate another test — candidates to remove so the suite stays fast and trusted.

Technical notes

Requirements are split into atomic rules (R-1, R-2 …), including rules implied by acceptance criteria. A rule counts as full only when its positive path and its main negative or boundary conditions are tested. Missing tests (MT-1, MT-2 …) name the requirement they close, with a type and a priority; redundant tests are those that map to no requirement or duplicate another.

References