diff --git a/features/steps/tdd_actor_run_response_steps.py b/features/steps/tdd_actor_run_response_steps.py
new file mode 100644
index 000000000..1c1f012cc
--- /dev/null
+++ b/features/steps/tdd_actor_run_response_steps.py
@@ -0,0 +1,70 @@
+"""Step definitions for TDD test: agents actor run returns no useful response.
+
+Issue: #10861 — agents actor run does not work
+TDD Issue: #10862
+
+This test captures the bug: when `agents actor run` is invoked with a built-in
+LLM actor name resolved from the registry, the command returns no useful
+response (either empty or an error).
+
+The @tdd_expected_fail tag inverts the result so CI passes while the bug exists.
+Once the fix is applied, the @tdd_expected_fail tag must be removed and the
+test must pass normally.
+"""
+
+from __future__ import annotations
+
+from typing import Any
+from unittest.mock import MagicMock, patch
+
+from behave import given, then, when
+from typer.testing import CliRunner
+
+from cleveragents.cli.commands.actor import app as actor_app
+
+
+@given("I have a mock LLM that returns {response}")
+def step_have_mock_llm(context: Any, response: str) -> None:
+    """Create a mock LLM that returns the specified response."""
+    context.mock_llm_response = response
+    context.mock_llm = MagicMock()
+    mock_response = MagicMock()
+    mock_response.content = response
+    context.mock_llm.invoke.return_value = mock_response
+
+
+@when("I run actor run with the built-in actor name and prompt {prompt}")
+def step_run_actor_run_with_builtin(context: Any, prompt: str) -> None:
+    """Invoke `agents actor run` with a built-in actor name and prompt."""
+    context.prompt = prompt
+
+    actor_name = next(
+        (name for name in context.actor_service.actors if "anthropic" in name.lower()),
+        None,
+    )
+    if not actor_name:
+        raise AssertionError("No built-in anthropic actor found")
+
+    context.actor_name = actor_name
+    context.runner = CliRunner()
+
+    def run_with_mocks():
+        with patch(
+            "cleveragents.providers.registry.ProviderRegistry.create_llm",
+            return_value=context.mock_llm,
+        ):
+            return context.runner.invoke(
+                actor_app,
+                ["run", actor_name, prompt],
+            )
+
+    context.result = run_with_mocks()
+
+
+@then("the actor run should return {expected}")
+def step_actor_run_should_return(context: Any, expected: str) -> None:
+    """Assert that the actor run command returned the expected response."""
+    output = context.result.output or ""
+    stderr = getattr(context.result, "stderr", "") or ""
+    combined = output + stderr
+    assert expected in combined, f"Expected '{expected}' in output: {combined!r}"
diff --git a/features/tdd_actor_run_response.feature b/features/tdd_actor_run_response.feature
new file mode 100644
index 000000000..da0fee385
--- /dev/null
+++ b/features/tdd_actor_run_response.feature
@@ -0,0 +1,12 @@
+@tdd_issue @tdd_issue_10861 @tdd_expected_fail
+Feature: TDD Issue #10862 — agents actor run returns no useful response
+  As a user who invokes `agents actor run` with a built-in LLM actor
+  I want to receive the LLM's response
+  So that the command is useful
+
+  Scenario: Actor run with built-in LLM actor returns the LLM response
+    Given a configured provider registry with anthropic/claude-sonnet-4-20250514
+    And the built-in actors have been ensured
+    And I have a mock LLM that returns "feep"
+    When I run actor run with the built-in actor name and prompt "ping"
+    Then the actor run should return "feep"
\ No newline at end of file