LiteLLM Append Trailing User Message

Claude 4.6+ does not support assistant message prefill.

Claude 4.6+ does not support assistant message prefill. Enable append_trailing_user_message to append a trailing user turn when the conversation ends with an assistant message (e.g. during reasoning).

append_trailing_user_message.py

"""
LiteLLM Append Trailing User Message
=====================================

Claude 4.6+ does not support assistant message prefill. Enable
`append_trailing_user_message` to append a trailing user turn when the
conversation ends with an assistant message (e.g. during reasoning).

Use `trailing_user_message_content` to customise the appended text (defaults to "continue").

Note: Claude 4.6+ models auto-detect and enable this flag automatically.
"""

from agno.agent import Agent
from agno.models.litellm import LiteLLM

# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------

agent = Agent(
    model=LiteLLM(
        id="anthropic/claude-sonnet-4-6",
        # Claude 4.6 rejects temperature + top_p together; drop top_p.
        top_p=None,
        append_trailing_user_message=True,
    ),
    reasoning=True,
    markdown=True,
)

# With custom trailing content
agent_custom = Agent(
    model=LiteLLM(
        id="anthropic/claude-sonnet-4-6",
        top_p=None,
        append_trailing_user_message=True,
        trailing_user_message_content="continue",
    ),
    reasoning=True,
    markdown=True,
)

# ---------------------------------------------------------------------------
# Run Agent
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    agent.print_response("What is 15 + 27?")
    agent_custom.print_response("What is 15 + 27?")

Run the Example

1. Set up your virtual environment

   Mac

   uv venv --python 3.12
   source .venv/bin/activate
   

   uv venv --python 3.12
   .venv\Scripts\activate
   
2. Install dependencies

   uv pip install -U agno litellm
   
3. Export your LiteLLM API key

   Mac/Linux

   export LITELLM_API_KEY="your_litellm_api_key_here"
   

   $Env:LITELLM_API_KEY="your_litellm_api_key_here"
   
4. Run the example

   Save the code above as append_trailing_user_message.py, then run:

   python append_trailing_user_message.py
   

Full source: cookbook/90_models/litellm/append_trailing_user_message.py