diff --git a/.gitignore b/.gitignore
index be4071bb..c3ff9440 100644
--- a/.gitignore
+++ b/.gitignore
@@ -37,3 +37,6 @@ ENV/
 .coverage
 htmlcov/
 .pytest_cache/
+
+# OSX
+.DS_Store
diff --git a/reasoning_gym/code/__init__.py b/reasoning_gym/code/__init__.py
new file mode 100644
index 00000000..19aca9df
--- /dev/null
+++ b/reasoning_gym/code/__init__.py
@@ -0,0 +1,7 @@
+"""
+Cognition tasks for training reasoning capabilities:
+- Code Analysis
+- Code Interpretation
+- Code Execution
+"""
+
diff --git a/reasoning_gym/code/bf.py b/reasoning_gym/code/bf.py
new file mode 100644
index 00000000..9f2f3381
--- /dev/null
+++ b/reasoning_gym/code/bf.py
@@ -0,0 +1,96 @@
+from dataclasses import dataclass
+from random import Random
+from typing import Dict, Optional
+
+import bfi
+from .contrib.bfit.Compiler import Compiler, Minify
+
+from ..data.wordle_words import wordle_words
+from ..factory import ProceduralDataset, register_dataset
+
+
+@dataclass
+class BFConfig:
+    """Configuration for BF task generation"""
+
+    seed: Optional[int] = None
+    size: int = 500
+
+
+class BFDataset(ProceduralDataset):
+    """Generates BF tasks"""
+
+    def __init__(self, config: BFConfig):
+        self._prompt_templates = [
+            "This is a BF (Brainf*ck) computer program. What is the output? \n\n{bf_program}",
+        ]
+        super().__init__(config=config, seed=config.seed, size=config.size)
+
+    def __getitem__(self, idx: int) -> dict:
+        """Generate a single BF task
+
+        Returns:
+            dict with keys:
+                - question: str, the task description with figlet string
+                - answer: str, the figlet encoded word
+                - metadata: dict with generation parameters
+        """
+        rng = Random(self.seed + idx)
+
+        bfit_code = self.generate_bfit_code(rng)
+        bf_program = self.compile_bfit_code_to_bf(bfit_code)
+
+        result = bfi.interpret(bf_program, buffer_output=True)
+
+        return {
+            "question": rng.choice(self._prompt_templates).format(bf_program=bf_program),
+            "answer": result,
+            "metadata": {"bfit_code": bfit_code, "bf_program": bf_program},
+        }
+
+    def generate_bfit_code(self, rng: Random) -> str:
+
+        bfit_template = """
+int main() {
+    int acc = 0;
+    int target = 15;
+    int x = 2;
+    int y = 3;
+    while (acc < target) {
+        acc = acc + x;
+        acc = acc + y;
+    }
+    printint(acc);
+}
+"""
+        rendered_bfit = bfit_template
+        return rendered_bfit
+
+    def compile_bfit_code_to_bf(self, bfit: str) -> str:
+        bf = Compiler.compile(bfit, optimize_code=True)
+        # bf = Minify.minify(bf) # Is this necessary?
+        return bf
+
+    def score_answer(self, answer: Optional[str], entry: Dict[str, any]) -> float:
+        """Determine if the solution provided solves the figlet task.
+
+        The function awards 1.0 for a correct answer and 0.1 points for each correct letter in the correct position,
+        with a maximum possible score of 1.0.
+
+        Args:
+            answer (Optional[str]): The user's answer.
+            entry (Dict[str, any]): The original dataset entry containing the correct answer.
+
+        Returns:
+            float: The computed score between 0.0 and 1.0.
+        """
+
+        if answer == None:
+            return 0.0
+        if answer != entry['answer']:
+            return 0.01
+        else:
+            return 1.0 # Yay
+
+# Register the dataset
+register_dataset("figlet_font", BFDataset, BFConfig)
diff --git a/tests/test_bf.py b/tests/test_bf.py
new file mode 100644
index 00000000..9340e9c4
--- /dev/null
+++ b/tests/test_bf.py
@@ -0,0 +1,37 @@
+import pytest
+
+from reasoning_gym.code.bf import BFConfig, BFDataset
+
+
+# def test_figlet_deterministic():
+#     """Test that dataset generates same items with same seed"""
+#     config = FigletFontConfig(seed=42, size=15)
+#     dataset1 = FigletFontDataset(config)
+#     dataset2 = FigletFontDataset(config)
+
+#     for i in range(15):  # Only check first 15 entries for speed
+#         assert dataset1[i] == dataset2[i]
+
+
+def test_bf():
+    """Test basic properties and solution of generated items"""
+    config = BFConfig(seed=42, size=40)
+    dataset = BFDataset(config)
+
+    for item in dataset:
+        assert isinstance(item, dict)
+        assert "question" in item
+        assert "answer" in item
+        assert "metadata" in item
+
+        # Check metadata contains required fields
+        assert "bfit_code" in item["metadata"]
+        assert "bf_program" in item["metadata"]
+
+        print(item["answer"])
+
+        # Test the scoring
+        assert dataset.score_answer(answer=item["answer"], entry=item) == 1.0
+        assert dataset.score_answer(answer=None, entry=item) == 0.0
+        assert dataset.score_answer(answer="Love is a battlefield", entry=item) == 0.01
+