mirror of
https://github.com/langgenius/dify-official-plugins.git
synced 2026-07-22 01:55:27 -04:00
Title: TongYi LLM Fails to Call MCP Functions with Array-Type Parameters in Dify Agent Template - "description" Error #785
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @DevElves on GitHub (Nov 7, 2025).
Self Checks
Dify version
1.9.2
Plugin version
0.1.0
Cloud or Self Hosted
Self Hosted (Docker)
Steps to reproduce
Problem Description:
When using the official Agent template in Dify to create an AI agent with MCP tools, TongYi LLM fails to call MCP functions that contain array-type parameters, throwing the error:
{"args":{"description":"[models] Error: 'description'"},"error_type":"InvokeError","message":"[models] Error: 'description'"}. This issue is specific to TongYi LLM and doesn't occur with other LLMs like DeepSeek which is also provided by the TongYi plugin. Additionally, MCP functions without array-type parameters work normally with TongYi LLM.Steps to Reproduce:
Create an AI agent using Dify's official Agent template
Add MCP tools that include functions with array-type parameters (e.g., List[str], Array[Any])
Configure TongYi LLM as the model provider
Attempt to call MCP functions with array parameters
Observe the "description" error in the response
Expected Behavior:
TongYi LLM should successfully call MCP functions with array-type parameters in Dify Agent template, similar to other LLMs like DeepSeek.
Actual Behavior:
TongYi LLM fails with the specific error:
json
✔️ Error log
{"args":{"description":"[models] Error: 'description'"},"error_type":"InvokeError","message":"[models] Error: 'description'"}
@dosubot[bot] commented on GitHub (Nov 23, 2025):
Hi, @DevElves. I'm Dosu, and I'm helping the dify-official-plugins team manage their backlog and am marking this issue as stale.
Issue Summary:
Next Steps:
Thank you for your understanding and contribution!