diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
new file mode 100644
index 0000000..7ccf9a5
--- /dev/null
+++ b/.github/workflows/test.yml
@@ -0,0 +1,19 @@
+name: Test
+on: [push, pull_request]
+
+jobs:
+  test:
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+        python-version: [3.6, 3.7, 3.8]
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Test
+        run: python -m unittest
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..58200d4
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+__pycache__/
diff --git a/README.md b/README.md
index 942f76f..1df2030 100644
--- a/README.md
+++ b/README.md
@@ -1,15 +1,64 @@
 # Codewars Test Framework for Python
 
+### Installation
 
-## Example
+```bash
+pip install git+https://github.com/codewars/python-test-framework.git#egg=codewars_test
+```
+
+### Basic Example
 
 ```python
-from solution import add
 import codewars_test as test
+from solution import add
 
 @test.describe('Example Tests')
 def example_tests():
+
     @test.it('Example Test Case')
     def example_test_case():
         test.assert_equals(add(1, 1), 2, 'Optional Message on Failure')
 ```
+
+<!--
+
+### Using Other Assertions
+
+> NOTE: This is not ready for production because the passed test case is not reported correctly.
+> See [#9](https://github.com/codewars/python-test-framework/issues/9).
+
+Any function that raises an `AssertionError` can be used instead of `codewars_test` assertions:
+
+```python
+import numpy as np
+import pandas as pd
+import codewars_test as test
+
+@test.describe('Example Tests')
+def test_custom_assertions():
+
+    @test.it('Test something in numpy')
+    def test_numpy_assertion():
+        actual = np.reshape(range(16), [4, 4])
+        expected = np.reshape(range(16, 0, -1), [4, 4])
+        np.testing.assert_equal(expected, actual)
+
+    @test.it('Test something in pandas')
+    def test_pandas_assertion():
+        actual = pd.DataFrame({'foo': [1, 2, 3]})
+        expected = pd.DataFrame({'foo': [1, 42, 3]})
+        pd.testing.assert_frame_equal(expected, actual)
+
+    @test.it('Test something using a custom assertion')
+    def test_custom_assertion():
+        def custom_assert_eq(actual, expected, msg=None):
+            if actual != expected:
+                default_msg = f'`{actual}` did not equal expected `{expected}`'
+                raise AssertionError(default_msg if msg is None else msg)
+
+        actual = 2
+        expected = 1
+        custom_assert_eq(actual, expected)
+```
+
+-->
diff --git a/codewars_test/test_framework.py b/codewars_test/test_framework.py
index 50b9f85..7b6fc6a 100644
--- a/codewars_test/test_framework.py
+++ b/codewars_test/test_framework.py
@@ -11,7 +11,8 @@ def format_message(message):
 
 def display(type, message, label="", mode=""):
     print("\n<{0}:{1}:{2}>{3}".format(
-        type.upper(), mode.upper(), label, format_message(message)))
+        type.upper(), mode.upper(), label, format_message(message))
+        , flush=True)
 
 
 def expect(passed=None, message=None, allow_raise=False):
@@ -51,8 +52,8 @@ def expect_error(message, function, exception=Exception):
         function()
     except exception:
         passed = True
-    except Exception:
-        pass
+    except Exception as e:
+        message = "{}: {} should be {}".format(message or "Unexpected exception", repr(e), repr(exception))
     expect(passed, message)
 
 
@@ -62,7 +63,7 @@ def expect_no_error(message, function, exception=BaseException):
     except exception as e:
         fail("{}: {}".format(message or "Unexpected exception", repr(e)))
         return
-    except Exception:
+    except:
         pass
     pass_()
 
diff --git a/gen-fixture.sh b/gen-fixture.sh
new file mode 100755
index 0000000..4093395
--- /dev/null
+++ b/gen-fixture.sh
@@ -0,0 +1,7 @@
+#!/bin/bash
+# ./gen-fixture.sh tests/fixtures/example.py
+# ./gen-fixture.sh tests/fixtures/example.py expected
+# ./gen-fixture.sh tests/fixtures/example.py sample
+# for f in $(ls tests/fixtures/*.py); do ./gen-fixture.sh "$f"; done
+
+PYTHONPATH=./ python "$1" > "${1%.py}.${2:-expected}.txt"
diff --git a/setup.py b/setup.py
index dc916f3..fea4d4d 100644
--- a/setup.py
+++ b/setup.py
@@ -2,10 +2,10 @@
 
 setup(
     name="codewars_test",
-    version="0.1.0",
+    version="0.2.1",
     packages=["codewars_test"],
     license="MIT",
     description="Codewars test framework for Python",
     install_requires=[],
-    url="https://github.com/Codewars/python-test-framework",
+    url="https://github.com/codewars/python-test-framework",
 )
diff --git a/tests/__init__.py b/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/fixtures/expect_error_sample.expected.txt b/tests/fixtures/expect_error_sample.expected.txt
new file mode 100644
index 0000000..b0cb56d
--- /dev/null
+++ b/tests/fixtures/expect_error_sample.expected.txt
@@ -0,0 +1,76 @@
+
+<DESCRIBE::>expect_error, new version
+
+<IT::>f0 raises nothing
+
+<FAILED::>f0 did not raise any exception
+
+<FAILED::>f0 did not raise Exception
+
+<FAILED::>f0 did not raise ArithmeticError
+
+<FAILED::>f0 did not raise ZeroDivisionError
+
+<FAILED::>f0 did not raise LookupError
+
+<FAILED::>f0 did not raise KeyError
+
+<FAILED::>f0 did not raise OSError
+
+<COMPLETEDIN::>0.03
+
+<IT::>f1 raises Exception
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<FAILED::>f1 did not raise ArithmeticError: Exception() should be <class 'ArithmeticError'>
+
+<FAILED::>f1 did not raise ZeroDivisionError: Exception() should be <class 'ZeroDivisionError'>
+
+<FAILED::>f1 did not raise LookupError: Exception() should be <class 'LookupError'>
+
+<FAILED::>f1 did not raise KeyError: Exception() should be <class 'KeyError'>
+
+<FAILED::>f1 did not raise OSError: Exception() should be <class 'OSError'>
+
+<COMPLETEDIN::>0.02
+
+<IT::>f2 raises Exception >> ArithmeticError >> ZeroDivisionError
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<FAILED::>f2 did not raise LookupError: ZeroDivisionError('integer division or modulo by zero') should be <class 'LookupError'>
+
+<FAILED::>f2 did not raise KeyError: ZeroDivisionError('integer division or modulo by zero') should be <class 'KeyError'>
+
+<FAILED::>f2 did not raise OSError: ZeroDivisionError('integer division or modulo by zero') should be <class 'OSError'>
+
+<COMPLETEDIN::>0.02
+
+<IT::>f3 raises Exception >> LookupError >> KeyError
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<FAILED::>f3 did not raise ArithmeticError: KeyError(1) should be <class 'ArithmeticError'>
+
+<FAILED::>f3 did not raise ZeroDivisionError: KeyError(1) should be <class 'ZeroDivisionError'>
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<FAILED::>f3 did not raise OSError: KeyError(1) should be <class 'OSError'>
+
+<COMPLETEDIN::>0.02
+
+<COMPLETEDIN::>0.11
\ No newline at end of file
diff --git a/tests/fixtures/expect_error_sample.py b/tests/fixtures/expect_error_sample.py
new file mode 100644
index 0000000..e76962f
--- /dev/null
+++ b/tests/fixtures/expect_error_sample.py
@@ -0,0 +1,60 @@
+# https://www.codewars.com/kumite/5ab735bee7093b17b2000084?sel=5ab735bee7093b17b2000084
+import codewars_test as test
+
+
+def f0():
+    pass
+
+
+# BaseException >> Exception
+def f1():
+    raise Exception()
+
+
+# BaseException >> Exception >> ArithmeticError >> ZeroDivisionError
+def f2():
+    return 1 // 0
+
+
+# BaseException >> Exception >> LookupError >> KeyError
+def f3():
+    return {}[1]
+
+
+excn = (
+    "Exception",
+    "ArithmeticError",
+    "ZeroDivisionError",
+    "LookupError",
+    "KeyError",
+    "OSError",
+)
+exc = (Exception, ArithmeticError, ZeroDivisionError, LookupError, KeyError, OSError)
+
+
+@test.describe("expect_error, new version")
+def d2():
+    @test.it("f0 raises nothing")
+    def i0():
+        test.expect_error("f0 did not raise any exception", f0)
+        for i in range(6):
+            test.expect_error("f0 did not raise {}".format(excn[i]), f0, exc[i])
+
+    @test.it("f1 raises Exception")
+    def i1():
+        test.expect_error("f1 did not raise Exception", f1)
+        for i in range(6):
+            test.expect_error("f1 did not raise {}".format(excn[i]), f1, exc[i])
+
+    @test.it("f2 raises Exception >> ArithmeticError >> ZeroDivisionError")
+    def i2():
+        test.expect_error("f2 did not raise Exception", f2)
+        for i in range(6):
+            test.expect_error("f2 did not raise {}".format(excn[i]), f2, exc[i])
+
+    @test.it("f3 raises Exception >> LookupError >> KeyError")
+    def i3():
+        test.expect_error("f3 did not raise Exception", f3)
+        for i in range(6):
+            test.expect_error("f3 did not raise {}".format(excn[i]), f3, exc[i])
+
diff --git a/tests/fixtures/multiple_groups.expected.txt b/tests/fixtures/multiple_groups.expected.txt
new file mode 100644
index 0000000..f9432cf
--- /dev/null
+++ b/tests/fixtures/multiple_groups.expected.txt
@@ -0,0 +1,20 @@
+
+<DESCRIBE::>group 1
+
+<IT::>test 1
+
+<FAILED::>1 should equal 2
+
+<COMPLETEDIN::>0.01
+
+<COMPLETEDIN::>0.02
+
+<DESCRIBE::>group 2
+
+<IT::>test 1
+
+<FAILED::>1 should equal 2
+
+<COMPLETEDIN::>0.00
+
+<COMPLETEDIN::>0.01
diff --git a/tests/fixtures/multiple_groups.py b/tests/fixtures/multiple_groups.py
new file mode 100644
index 0000000..0a24b59
--- /dev/null
+++ b/tests/fixtures/multiple_groups.py
@@ -0,0 +1,15 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    @test.it("test 1")
+    def test_1():
+        test.assert_equals(1, 2)
+
+
+@test.describe("group 2")
+def group_2():
+    @test.it("test 1")
+    def test_1():
+        test.assert_equals(1, 2)
diff --git a/tests/fixtures/nested_groups.expected.txt b/tests/fixtures/nested_groups.expected.txt
new file mode 100644
index 0000000..7445137
--- /dev/null
+++ b/tests/fixtures/nested_groups.expected.txt
@@ -0,0 +1,14 @@
+
+<DESCRIBE::>group 1
+
+<DESCRIBE::>group 1 1
+
+<IT::>test 1
+
+<FAILED::>1 should equal 2
+
+<COMPLETEDIN::>0.01
+
+<COMPLETEDIN::>0.02
+
+<COMPLETEDIN::>0.02
diff --git a/tests/fixtures/nested_groups.py b/tests/fixtures/nested_groups.py
new file mode 100644
index 0000000..b9e3636
--- /dev/null
+++ b/tests/fixtures/nested_groups.py
@@ -0,0 +1,10 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    @test.describe("group 1 1")
+    def group_1_1():
+        @test.it("test 1")
+        def test_1():
+            test.assert_equals(1, 2)
diff --git a/tests/fixtures/old_group.expected.txt b/tests/fixtures/old_group.expected.txt
new file mode 100644
index 0000000..e450a9b
--- /dev/null
+++ b/tests/fixtures/old_group.expected.txt
@@ -0,0 +1,6 @@
+
+<DESCRIBE::>group 1
+
+<IT::>test 1
+
+<PASSED::>Test Passed
diff --git a/tests/fixtures/old_group.py b/tests/fixtures/old_group.py
new file mode 100644
index 0000000..db08a36
--- /dev/null
+++ b/tests/fixtures/old_group.py
@@ -0,0 +1,6 @@
+# Deprecated and should not be used
+import codewars_test as test
+
+test.describe("group 1")
+test.it("test 1")
+test.assert_equals(1, 1)
diff --git a/tests/fixtures/old_top_level_assertion_fail.expected.txt b/tests/fixtures/old_top_level_assertion_fail.expected.txt
new file mode 100644
index 0000000..0475c84
--- /dev/null
+++ b/tests/fixtures/old_top_level_assertion_fail.expected.txt
@@ -0,0 +1,2 @@
+
+<FAILED::>1 should equal 2
diff --git a/tests/fixtures/old_top_level_assertion_fail.py b/tests/fixtures/old_top_level_assertion_fail.py
new file mode 100644
index 0000000..c801dc5
--- /dev/null
+++ b/tests/fixtures/old_top_level_assertion_fail.py
@@ -0,0 +1,4 @@
+# Deprecated and should not be used
+import codewars_test as test
+
+test.assert_equals(1, 2)
diff --git a/tests/fixtures/old_top_level_assertion_fail_pass.expected.txt b/tests/fixtures/old_top_level_assertion_fail_pass.expected.txt
new file mode 100644
index 0000000..d47bd4f
--- /dev/null
+++ b/tests/fixtures/old_top_level_assertion_fail_pass.expected.txt
@@ -0,0 +1,4 @@
+
+<FAILED::>1 should equal 2
+
+<PASSED::>Test Passed
diff --git a/tests/fixtures/old_top_level_assertion_fail_pass.py b/tests/fixtures/old_top_level_assertion_fail_pass.py
new file mode 100644
index 0000000..690bfc5
--- /dev/null
+++ b/tests/fixtures/old_top_level_assertion_fail_pass.py
@@ -0,0 +1,4 @@
+import codewars_test as test
+
+test.assert_equals(1, 2)
+test.assert_equals(1, 1)
diff --git a/tests/fixtures/old_top_level_assertion_pass.expected.txt b/tests/fixtures/old_top_level_assertion_pass.expected.txt
new file mode 100644
index 0000000..4bfed57
--- /dev/null
+++ b/tests/fixtures/old_top_level_assertion_pass.expected.txt
@@ -0,0 +1,2 @@
+
+<PASSED::>Test Passed
diff --git a/tests/fixtures/old_top_level_assertion_pass.py b/tests/fixtures/old_top_level_assertion_pass.py
new file mode 100644
index 0000000..3a814f8
--- /dev/null
+++ b/tests/fixtures/old_top_level_assertion_pass.py
@@ -0,0 +1,4 @@
+# Deprecated and should not be used
+import codewars_test as test
+
+test.assert_equals(1, 1)
diff --git a/tests/fixtures/passing_failing.expected.txt b/tests/fixtures/passing_failing.expected.txt
new file mode 100644
index 0000000..edeea1f
--- /dev/null
+++ b/tests/fixtures/passing_failing.expected.txt
@@ -0,0 +1,16 @@
+
+<DESCRIBE::>group 1
+
+<IT::>test 1
+
+<PASSED::>Test Passed
+
+<COMPLETEDIN::>0.00
+
+<IT::>test 2
+
+<FAILED::>1 should equal 2
+
+<COMPLETEDIN::>0.00
+
+<COMPLETEDIN::>0.02
diff --git a/tests/fixtures/passing_failing.py b/tests/fixtures/passing_failing.py
new file mode 100644
index 0000000..1d76fc4
--- /dev/null
+++ b/tests/fixtures/passing_failing.py
@@ -0,0 +1,12 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    @test.it("test 1")
+    def test_1():
+        test.assert_equals(1, 1)
+
+    @test.it("test 2")
+    def test_2():
+        test.assert_equals(1, 2)
diff --git a/tests/fixtures/single_failing.expected.txt b/tests/fixtures/single_failing.expected.txt
new file mode 100644
index 0000000..f81b2a3
--- /dev/null
+++ b/tests/fixtures/single_failing.expected.txt
@@ -0,0 +1,10 @@
+
+<DESCRIBE::>group 1
+
+<IT::>test 1
+
+<FAILED::>1 should equal 2
+
+<COMPLETEDIN::>0.00
+
+<COMPLETEDIN::>0.01
diff --git a/tests/fixtures/single_failing.py b/tests/fixtures/single_failing.py
new file mode 100644
index 0000000..8580ff8
--- /dev/null
+++ b/tests/fixtures/single_failing.py
@@ -0,0 +1,8 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    @test.it("test 1")
+    def test_1():
+        test.assert_equals(1, 2)
diff --git a/tests/fixtures/single_passing.expected.txt b/tests/fixtures/single_passing.expected.txt
new file mode 100644
index 0000000..b35ba0e
--- /dev/null
+++ b/tests/fixtures/single_passing.expected.txt
@@ -0,0 +1,10 @@
+
+<DESCRIBE::>group 1
+
+<IT::>test 1
+
+<PASSED::>Test Passed
+
+<COMPLETEDIN::>0.01
+
+<COMPLETEDIN::>0.02
diff --git a/tests/fixtures/single_passing.py b/tests/fixtures/single_passing.py
new file mode 100644
index 0000000..626fea7
--- /dev/null
+++ b/tests/fixtures/single_passing.py
@@ -0,0 +1,8 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    @test.it("test 1")
+    def test_1():
+        test.assert_equals(1, 1)
diff --git a/tests/fixtures/timeout_failing.expected.txt b/tests/fixtures/timeout_failing.expected.txt
new file mode 100644
index 0000000..cf478d6
--- /dev/null
+++ b/tests/fixtures/timeout_failing.expected.txt
@@ -0,0 +1,6 @@
+
+<DESCRIBE::>group 1
+
+<FAILED::>Exceeded time limit of 0.010 seconds
+
+<COMPLETEDIN::>30.41
diff --git a/tests/fixtures/timeout_failing.py b/tests/fixtures/timeout_failing.py
new file mode 100644
index 0000000..00c2d06
--- /dev/null
+++ b/tests/fixtures/timeout_failing.py
@@ -0,0 +1,11 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    @test.timeout(0.01)
+    def test_1():
+        x = 0
+        while x < 10 ** 9:
+            x += 1
+        test.pass_()
diff --git a/tests/fixtures/timeout_passing.expected.txt b/tests/fixtures/timeout_passing.expected.txt
new file mode 100644
index 0000000..a0894ac
--- /dev/null
+++ b/tests/fixtures/timeout_passing.expected.txt
@@ -0,0 +1,8 @@
+
+<DESCRIBE::>group 1
+
+<PASSED::>Test Passed
+
+<PASSED::>Test Passed
+
+<COMPLETEDIN::>17.19
diff --git a/tests/fixtures/timeout_passing.py b/tests/fixtures/timeout_passing.py
new file mode 100644
index 0000000..aa63d1c
--- /dev/null
+++ b/tests/fixtures/timeout_passing.py
@@ -0,0 +1,10 @@
+import codewars_test as test
+
+
+@test.describe("group 1")
+def group_1():
+    # This outputs 2 PASSED
+    @test.timeout(0.01)
+    def test_1():
+        test.assert_equals(1, 1)
+
diff --git a/tests/test_outputs.py b/tests/test_outputs.py
new file mode 100644
index 0000000..e35f8a2
--- /dev/null
+++ b/tests/test_outputs.py
@@ -0,0 +1,81 @@
+import unittest
+import subprocess
+import os
+import re
+from pathlib import Path
+
+
+class TestOutputs(unittest.TestCase):
+    pass
+
+
+def test_against_expected(test_file, expected_file, env):
+    def test(self):
+        # Using `stdout=PIPE, stderr=PIPE` for Python 3.6 compatibility instead of `capture_output=True`
+        result = subprocess.run(
+            ["python", test_file],
+            env=env,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+        )
+        with open(expected_file, "r", encoding="utf-8") as r:
+            # Allow duration to change
+            expected = re.sub(r"([()])", r"\\\1", r.read())
+            expected = re.sub(
+                r"(?<=<COMPLETEDIN::>)\d+(?:\.\d+)?", r"\\d+(?:\\.\\d+)?", expected
+            )
+
+            self.assertRegex(result.stdout.decode("utf-8"), expected)
+
+    return test
+
+
+def get_commands(output):
+    return re.findall(r"<(?:DESCRIBE|IT|PASSED|FAILED|ERROR|COMPLETEDIN)::>", output)
+
+
+def test_against_sample(test_file, sample_file, env):
+    def test(self):
+        # Using `stdout=PIPE, stderr=PIPE` for Python 3.6 compatibility instead of `capture_output=True`
+        result = subprocess.run(
+            ["python", test_file],
+            env=env,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+        )
+        with open(sample_file, "r", encoding="utf-8") as r:
+            # Ensure that it contains the same output structure
+            self.assertEqual(
+                get_commands(result.stdout.decode("utf-8")), get_commands(r.read())
+            )
+
+    return test
+
+
+def define_tests():
+    fixtures_dir = os.path.join(os.path.dirname(os.path.realpath(__file__)), "fixtures")
+    package_dir = Path(fixtures_dir).parent.parent
+    files = (f for f in os.listdir(fixtures_dir) if f.endswith(".py"))
+    for f in files:
+        expected_file = os.path.join(fixtures_dir, f.replace(".py", ".expected.txt"))
+        if os.path.exists(expected_file):
+            test_func = test_against_expected(
+                os.path.join(fixtures_dir, f),
+                expected_file,
+                {"PYTHONPATH": package_dir},
+            )
+        else:
+            # Use `.sample.txt` when testing against outputs with more variables.
+            # This version only checks for the basic structure.
+            test_func = test_against_sample(
+                os.path.join(fixtures_dir, f),
+                os.path.join(fixtures_dir, f.replace(".py", ".sample.txt")),
+                {"PYTHONPATH": package_dir},
+            )
+        setattr(TestOutputs, "test_{0}".format(f.replace(".py", "")), test_func)
+
+
+define_tests()
+
+if __name__ == "__main__":
+    unittest.main()