Automatically identify potential boundary and exception cases from requirements, specifications, or existing code, and generate comprehensive test cases targeting boundary conditions, edge cases, and uncommon scenarios. Use this skill when analyzing programs, code repositories, functions, or APIs to discover and test corner cases, null handling, overflow conditions, empty inputs, concurrent access patterns, and other exceptional scenarios that are often missed in standard testing.
93
92%
Does it follow best practices?
Impact
96%
1.03xAverage score across 3 eval scenarios
Passed
No known issues
Systematically identify boundary conditions, edge cases, and exceptional scenarios, then generate comprehensive tests to validate software behavior under uncommon conditions.
Analyze code, specifications, or requirements to identify:
Generate tests in multiple formats:
Support major programming languages:
Analyze each input parameter or data structure:
For numeric inputs:
- Minimum value (e.g., INT_MIN, 0, -∞)
- Maximum value (e.g., INT_MAX, 255, +∞)
- Zero
- One (unit value)
- Negative one
- Just below/above boundaries
- Overflow/underflow valuesFor collections (arrays, lists, sets):
- Empty collection
- Single element
- Two elements (minimal interaction)
- Maximum size (if bounded)
- All identical elements
- All unique elements
- Sorted vs unsorted
- Contains duplicatesFor strings:
- Empty string ("")
- Single character
- Very long string (10K+ chars)
- Unicode characters (emoji, RTL, special)
- Whitespace only (" ")
- Null terminator issues
- Special characters (\n, \t, \0)For pointers/references:
- Null/None/undefined
- Dangling pointer
- Self-reference (circular)
- UninitializedObject state:
- Uninitialized state
- Partially initialized
- Valid operational state
- Invalid/corrupted state
- Locked/busy state
- Disposed/freed statePrecondition violations:
- Missing required parameters
- Parameters in wrong order
- Invalid combinations
- Violated invariantsSuccess cases:
- Typical successful execution
- Boundary successful case
- Empty result (valid but empty)Failure cases:
- Expected exceptions/errors
- Timeout scenarios
- Partial failures
- Resource exhaustionTemporal patterns:
- First operation
- Last operation
- Repeated operations
- Alternating operations
- Concurrent operationsDependency patterns:
- Missing dependencies
- Circular dependencies
- Incompatible versions
- Network failuresFor each identified edge case, generate:
# Example: Testing a function that calculates factorial
def test_factorial_edge_cases():
# Zero (boundary)
assert factorial(0) == 1
# One (boundary)
assert factorial(1) == 1
# Negative (invalid input)
with pytest.raises(ValueError):
factorial(-1)
# Large value (overflow risk)
with pytest.raises(OverflowError):
factorial(10000)
# Non-integer (type boundary)
with pytest.raises(TypeError):
factorial(3.5)// Example: Testing an array processing function
describe('processArray edge cases', () => {
test('empty array', () => {
expect(processArray([])).toEqual([]);
});
test('single element', () => {
expect(processArray([1])).toEqual([1]);
});
test('null input', () => {
expect(() => processArray(null)).toThrow(TypeError);
});
test('undefined input', () => {
expect(() => processArray(undefined)).toThrow(TypeError);
});
test('all identical elements', () => {
expect(processArray([5, 5, 5, 5])).toEqual([5, 5, 5, 5]);
});
test('very large array', () => {
const largeArray = new Array(1000000).fill(1);
expect(() => processArray(largeArray)).not.toThrow();
});
});// Example: Testing string validation
@Test
public void testValidateString_EdgeCases() {
// Empty string
assertThrows(ValidationException.class, () -> validateString(""));
// Null
assertThrows(NullPointerException.class, () -> validateString(null));
// Single character
assertTrue(validateString("a"));
// Whitespace only
assertFalse(validateString(" "));
// Special characters
assertTrue(validateString("hello@world!"));
// Unicode
assertTrue(validateString("你好世界"));
// Very long string
String longString = "a".repeat(100000);
assertDoesNotThrow(() -> validateString(longString));
// Control characters
assertFalse(validateString("hello\0world"));
}// Example: Testing pointer operations
void test_pointer_edge_cases() {
// Null pointer
assert(safe_strlen(NULL) == -1);
// Empty string
assert(safe_strlen("") == 0);
// Single character
assert(safe_strlen("a") == 1);
// Very long string
char long_str[10000];
memset(long_str, 'a', 9999);
long_str[9999] = '\0';
assert(safe_strlen(long_str) == 9999);
}// Example: Testing concurrent access
func TestConcurrentAccess(t *testing.T) {
cache := NewCache()
// Race condition: multiple concurrent writes
t.Run("concurrent writes", func(t *testing.T) {
var wg sync.WaitGroup
for i := 0; i < 100; i++ {
wg.Add(1)
go func(val int) {
defer wg.Done()
cache.Set("key", val)
}(i)
}
wg.Wait()
// Verify cache is in valid state
_, ok := cache.Get("key")
assert.True(t, ok)
})
// Deadlock scenario
t.Run("potential deadlock", func(t *testing.T) {
done := make(chan bool)
go func() {
cache.Set("key1", 1)
cache.Set("key2", 2)
done <- true
}()
select {
case <-done:
// Success
case <-time.After(5 * time.Second):
t.Fatal("Deadlock detected")
}
})
}# Example: Testing state machine
def test_state_machine_edge_cases():
machine = StateMachine()
# Invalid initial transition
with pytest.raises(InvalidStateError):
machine.stop() # Can't stop before starting
# Double start
machine.start()
with pytest.raises(InvalidStateError):
machine.start() # Already started
# Transition from error state
machine.trigger_error()
with pytest.raises(InvalidStateError):
machine.resume() # Can't resume from error
# Valid state sequence
machine.reset()
machine.start()
machine.pause()
machine.resume()
machine.stop()
assert machine.state == "stopped"# Example: Testing memory and resource limits
def test_resource_limits():
# Memory exhaustion
with pytest.raises(MemoryError):
allocate_huge_buffer(size=10**15)
# File handle exhaustion
handles = []
try:
for i in range(10000):
handles.append(open(f'/tmp/test_{i}', 'w'))
except OSError as e:
assert "Too many open files" in str(e)
finally:
for h in handles:
h.close()
# Timeout scenarios
with pytest.raises(TimeoutError):
slow_operation(timeout=0.001) # Very short timeout
# Disk full simulation
with patch('os.write', side_effect=OSError("No space left on device")):
with pytest.raises(OSError):
write_large_file()Divide input domain into equivalence classes:
# For a function that accepts age (0-120)
# Partitions: Invalid (<0), Valid (0-120), Invalid (>120)
def test_age_validation():
# Invalid: negative
assert not is_valid_age(-1)
assert not is_valid_age(-100)
# Valid: boundaries
assert is_valid_age(0)
assert is_valid_age(120)
# Valid: typical
assert is_valid_age(25)
assert is_valid_age(65)
# Invalid: too large
assert not is_valid_age(121)
assert not is_valid_age(1000)Test at and around boundaries:
// For array index access (0 to length-1)
describe('array access boundaries', () => {
const arr = [10, 20, 30, 40, 50];
test('before first element', () => {
expect(() => arr.at(-1)).toThrow(RangeError);
});
test('first element (boundary)', () => {
expect(arr.at(0)).toBe(10);
});
test('second element (just inside)', () => {
expect(arr.at(1)).toBe(20);
});
test('last element (boundary)', () => {
expect(arr.at(4)).toBe(50);
});
test('after last element', () => {
expect(() => arr.at(5)).toThrow(RangeError);
});
});Generate random edge cases automatically:
from hypothesis import given, strategies as st
@given(st.lists(st.integers()))
def test_sort_properties(input_list):
"""Property: sorted list length equals input length"""
result = sort(input_list)
assert len(result) == len(input_list)
@given(st.lists(st.integers(), min_size=1))
def test_max_element_property(input_list):
"""Property: max element >= all elements"""
max_val = find_max(input_list)
assert all(max_val >= x for x in input_list)
@given(st.text())
def test_string_reverse_property(s):
"""Property: reversing twice returns original"""
assert reverse(reverse(s)) == sTest that tests actually catch bugs:
# Original function
def is_even(n):
return n % 2 == 0
# Mutant 1: Changed == to !=
def is_even_mutant1(n):
return n % 2 != 0 # Should be caught by tests
# Test that catches this mutation
def test_is_even():
assert is_even(0) == True # Catches mutant1
assert is_even(1) == False # Catches mutant1
assert is_even(2) == True # Catches mutant1Use this checklist when analyzing code:
For language-specific edge cases and test patterns, see the reference files:
Given a function specification:
def binary_search(arr: list[int], target: int) -> int:
"""
Find target in sorted array using binary search.
Returns index if found, -1 otherwise.
Requires: arr is sorted in ascending order
"""Generated edge case tests:
import pytest
class TestBinarySearchEdgeCases:
"""Comprehensive edge case tests for binary_search"""
# Collection boundaries
def test_empty_array(self):
"""Edge: Empty collection"""
assert binary_search([], 5) == -1
def test_single_element_found(self):
"""Edge: Single element - target present"""
assert binary_search([5], 5) == 0
def test_single_element_not_found(self):
"""Edge: Single element - target absent"""
assert binary_search([5], 3) == -1
def test_two_elements_first(self):
"""Edge: Minimal interaction - target is first"""
assert binary_search([1, 2], 1) == 0
def test_two_elements_second(self):
"""Edge: Minimal interaction - target is second"""
assert binary_search([1, 2], 2) == 1
# Position boundaries
def test_target_at_start(self):
"""Edge: Target at first position"""
assert binary_search([1, 2, 3, 4, 5], 1) == 0
def test_target_at_end(self):
"""Edge: Target at last position"""
assert binary_search([1, 2, 3, 4, 5], 5) == 4
def test_target_before_range(self):
"""Edge: Target less than all elements"""
assert binary_search([5, 10, 15], 3) == -1
def test_target_after_range(self):
"""Edge: Target greater than all elements"""
assert binary_search([5, 10, 15], 20) == -1
# Duplicate elements
def test_all_elements_same(self):
"""Edge: All elements identical"""
arr = [5, 5, 5, 5, 5]
assert binary_search(arr, 5) in range(len(arr))
def test_duplicates_present(self):
"""Edge: Target appears multiple times"""
# Should return any valid index
result = binary_search([1, 2, 2, 2, 3], 2)
assert result in [1, 2, 3]
# Numeric boundaries
def test_negative_values(self):
"""Edge: Negative numbers"""
assert binary_search([-10, -5, 0, 5, 10], -5) == 1
def test_zero_value(self):
"""Edge: Zero in array"""
assert binary_search([-2, -1, 0, 1, 2], 0) == 2
def test_large_values(self):
"""Edge: Near integer limits"""
import sys
max_int = sys.maxsize
assert binary_search([0, max_int], max_int) == 1
# Size boundaries
def test_very_large_array(self):
"""Edge: Performance with large array"""
arr = list(range(1000000))
assert binary_search(arr, 999999) == 999999
# Invalid inputs
def test_null_array(self):
"""Edge: Null/None array"""
with pytest.raises(TypeError):
binary_search(None, 5)
def test_unsorted_array(self):
"""Edge: Precondition violation - unsorted array"""
# May return incorrect result or -1
# This tests documentation compliance
result = binary_search([5, 1, 3, 2, 4], 3)
# Behavior is undefined, just ensure no crash
assert isinstance(result, int)This comprehensive test suite covers 18 distinct edge cases across multiple categories.
0f00a4f
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.