Attaching Composio tool (i.e filetool_write) causes non stop warning message which ends up filling the context window and floods server log

[a8nova]

Describe the bug
When attaching a Composio tool like filetool_write, there is a warning that gets dumped from https://github.com/letta-ai/letta/blob/main/letta/local_llm/utils.py#L130 and there is additional messages which i think are coming from the LLM backend being used when tool calling, look at the 1st screenshot for the nonstop messages Please provide a value of type string.

/app/letta/local_llm/utils.py:130: UserWarning: num_tokens_from_functions: Unsupported field default in function {'name': 'filetool_write', 'description': 'Write The Given Content To A File.\n Note:     This Action Will Replace The Existing Content In The The File, And Completely Rewrite The File, If You Want To Edit A Specific Portion Of The File Use `Edit` Tool Instead.', 'parameters': {'type': 'object', 'properties': {'file_manager_id': {'type': 'string', 'description': 'ID of the file manager where the file will be opened, if not provided the recent file manager will be used to execute the action. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string. Please provide a value of type string

Above is just a warning but floods the server logs since it gets printed nonstop and also fills the context window. It looks like the actual message that floods the server logs and fills the context window is coming from somewhere else (not from the open source framework). Below screenshot of my terminal shows the repeated messages.

Please describe your setup

• How did you install letta?
  • I installed with Docker, I followed https://docs.letta.com/guides/server/docker
• Describe your setup
  • MacOS with Apple chip
  • I am running letta via:

docker run \
  -v ~/.letta/.persist/pgdata:/var/lib/postgresql/data \
  -p 8283:8283 \
  --env-file .env \
  letta/letta:latest

Screenshots
If applicable, add screenshots to help explain your problem.

Additional context
Add any other context about the problem here.

Letta Config
Please attach your ~/.letta/config file or copy paste it below.

[defaults]
preset = memgpt_chat
persona = sam_pov
human = basic

[archival_storage]
type = postgres
path = /Users/alazarshenkute/.letta
uri = postgresql://postgres:@/postgres?host=/Users/alazarshenkute/.letta/data

[recall_storage]
type = postgres
path = /Users/alazarshenkute/.letta
uri = postgresql://postgres:@/postgres?host=/Users/alazarshenkute/.letta/data

[metadata_storage]
type = sqlite
path = /Users/alazarshenkute/.letta

[version]
letta_version = 0.6.13

---

If you're not using OpenAI, please provide additional information on your local LLM setup:

Local LLM details

If you are trying to run Letta with local LLMs, please provide the following information:

• The exact model you're trying to use (e.g. dolphin-2.1-mistral-7b.Q6_K.gguf)
• The local LLM backend you are using (web UI? LM Studio?)
• Your hardware for the local LLM backend (local computer? operating system? remote RunPod?)

[mattzh72]

Digging into this actively!