diff --git a/autogpt_platform/backend/backend/copilot/sdk/agent_generation_guide.md b/autogpt_platform/backend/backend/copilot/sdk/agent_generation_guide.md
index 00750e3b1f..efa3ebeed3 100644
--- a/autogpt_platform/backend/backend/copilot/sdk/agent_generation_guide.md
+++ b/autogpt_platform/backend/backend/copilot/sdk/agent_generation_guide.md
@@ -37,7 +37,7 @@ Before running the workflow below, ALWAYS decompose the goal first:
 4. Only after approval, continue with "Workflow for Creating/Editing Agents".
 
 For simple goals (1-2 blocks), keep steps brief (2-3 steps).
-For complex goals, decompose into 4-8 steps max.
+For complex goals, use as many steps as needed.
 
 ### Workflow for Creating/Editing Agents
 
diff --git a/autogpt_platform/backend/backend/copilot/tools/decompose_goal.py b/autogpt_platform/backend/backend/copilot/tools/decompose_goal.py
index 925336a9a3..36d7f1d4f9 100644
--- a/autogpt_platform/backend/backend/copilot/tools/decompose_goal.py
+++ b/autogpt_platform/backend/backend/copilot/tools/decompose_goal.py
@@ -18,8 +18,6 @@ from .models import (
 
 logger = logging.getLogger(__name__)
 
-# Matches the guide's "4-8 steps max" constraint.
-MAX_STEPS = 8
 DEFAULT_ACTION = "add_block"
 VALID_ACTIONS = {"add_block", "connect_blocks", "configure", "add_input", "add_output"}
 
@@ -275,13 +273,6 @@ class DecomposeGoalTool(BaseTool):
                 session_id=session_id,
             )
 
-        if len(steps) > MAX_STEPS:
-            return ErrorResponse(
-                message=f"Too many steps ({len(steps)}). Keep the plan to {MAX_STEPS} steps max.",
-                error="too_many_steps",
-                session_id=session_id,
-            )
-
         decomposition_steps: list[DecompositionStepModel] = []
         for i, step in enumerate(steps):
             if not isinstance(step, dict):
diff --git a/autogpt_platform/backend/backend/copilot/tools/decompose_goal_test.py b/autogpt_platform/backend/backend/copilot/tools/decompose_goal_test.py
index e234aa3a5f..51d99d0434 100644
--- a/autogpt_platform/backend/backend/copilot/tools/decompose_goal_test.py
+++ b/autogpt_platform/backend/backend/copilot/tools/decompose_goal_test.py
@@ -13,7 +13,6 @@ from ._test_data import make_session
 from .decompose_goal import (
     AUTO_APPROVE_CLIENT_SECONDS,
     DEFAULT_ACTION,
-    MAX_STEPS,
     DecomposeGoalTool,
     _no_user_action_since,
     cancel_auto_approve,
@@ -190,38 +189,6 @@ async def test_empty_steps_returns_error(tool: DecomposeGoalTool, session):
     assert result.error == "missing_steps"
 
 
-@pytest.mark.asyncio
-async def test_too_many_steps_returns_error(tool: DecomposeGoalTool, session):
-    many_steps = [
-        {"description": f"Step {i}", "action": "add_block"}
-        for i in range(MAX_STEPS + 1)
-    ]
-    result = await tool._execute(
-        user_id=_USER_ID,
-        session=session,
-        goal="Over-engineered agent",
-        steps=many_steps,
-    )
-    assert isinstance(result, ErrorResponse)
-    assert result.error == "too_many_steps"
-
-
-@pytest.mark.asyncio
-async def test_exactly_max_steps_succeeds(tool: DecomposeGoalTool, session):
-    """Exactly MAX_STEPS steps should succeed."""
-    max_steps = [
-        {"description": f"Step {i}", "action": "add_block"} for i in range(MAX_STEPS)
-    ]
-    result = await tool._execute(
-        user_id=_USER_ID,
-        session=session,
-        goal="Complex agent",
-        steps=max_steps,
-    )
-    assert isinstance(result, TaskDecompositionResponse)
-    assert len(result.steps) == MAX_STEPS
-
-
 # ---------------------------------------------------------------------------
 # Validation — malformed step items
 # ---------------------------------------------------------------------------