Dify1.1.2，local docker deploy, issue:agent node function calling problem

[lazywking]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• Please do not modify this template :) and fill in all the required fields.

Content

2025-04-09 07:34:47.562 WARNING [Thread-1059 (_generate_worker)] [provider_configuration.py:989] - get custom model schema failed, {"code":0,"message":"success","data":{"model_schema":null}}
2025-04-09 07:34:47.580 DEBUG [Thread-1059 (_generate_worker)] [connectionpool.py:241] - Starting new HTTP connection (1): plugin_daemon:5002
2025-04-09 07:34:47.584 DEBUG [Thread-1059 (_generate_worker)] [connectionpool.py:544] - http://plugin_daemon:5002 "GET /plugin/9e3c5683-bc96-4e2d-b67f-34be8a56929f/management/tools?page=1&page_size=256 HTTP/1.1" 200 None

[AgentLogEvent]
2025-04-09 07:34:48.693 DEBUG [Thread-1060 (process_request_thread)] [connectionpool.py:544] - http://plugin_daemon:5002 "POST /plugin/9e3c5683-bc96-4e2d-b67f-34be8a56929f/dispatch/llm/invoke HTTP/1.1" 200 None
2025-04-09 07:34:48.698 INFO [Thread-1060 (process_request_thread)] [_internal.py:97] - 172.18.0.7 - - [09/Apr/2025 07:34:48] "POST /inner/api/invoke/llm HTTP/1.1" 200 -

[on_llm_after_invoke]
Content:
Model: qwen-max
Usage: prompt_tokens=279 prompt_unit_price=Decimal('0.02') prompt_price_unit=Decimal('0.001') prompt_price=Decimal('0.00558') completion_tokens=17 completion_unit_price=Decimal('0.06') completion_price_unit=Decimal('0.001') completion_price=Decimal('0.00102') total_tokens=296 total_price=Decimal('0.0066') currency='RMB' latency=1.1568843959830701
System Fingerprint: None

[NodeRunFailedEvent]
Node ID: 1742202983138
Node Title: 账号Agent
Type: agent
Error: Failed to transform agent message: PluginInvokeError: {"args":{},"error_type":"JSONDecodeError","message":"Expecting ':' delimiter: line 1 column 13 (char 12)"}
Inputs: {'model': {'provider': 'langgenius/tongyi/tongyi', 'model': 'qwen-max', 'model_type': 'llm', 'mode': 'chat', 'completion_params': {}, 'type': 'model-selector'}, 'tools': [{'provider_name': '71b1f4d6-e6cc-4b3e-8036-6551029f176e', 'type': 'workflow', 'tool_name': 'usernameCheck', 'tool_label': 'UserNameCheck', 'settings': {}, 'parameters': {'username': None}, 'enabled': True, 'extra': {'description': ''}}], 'instruction': '你是有用的助手\nhistory中是你和玩家的历史聊天记录\n你的任务是处理账号相关的问题，其他问题不需要处理\n从用户查询中提取用户名（通常跟在“我的用户名是”之后），并调用 checkUserAccount 工具\n工具调用参数必须是严格的 JSON 格式，使用双引号，例如我的用户名是wa请查看账号是否正常得到 {"username": "wa"}\n如果无法提取用户名，返回 {"username": null} 并说明原因', 'query': "\n['用户: 我的用户名是wa请查看账号是否正常']\n"}
Process Data:
Outputs:

[GraphRunFailedEvent] reason: Failed to transform agent message: PluginInvokeError: {"args":{},"error_type":"JSONDecodeError","message":"Expecting ':' delimiter: line 1 column 13 (char 12)"}
2025-04-09 07:34:48.748 DEBUG [Thread-1058 (process_request_thread)] [based_generate_task_pipeline.py:45] - error: Run failed: Failed to transform agent message: PluginInvokeError: {"args":{},"error_type":"JSONDecodeError","message":"Expecting ':' delimiter: line 1 column 13 (char 12)"}
2025-04-09 07:35:01.151 DEBUG [Thread-1062 (process_request_thread)] [connectionpool.py:241] - Starting new HTTP connection (1): plugin_daemon:5002

does anyone have same issue?

[dosubot[bot]]

I found 1 similar issue that might be helpful:

• docker-plugin_daemon-1 plugin error title - No solution was found in the comments.

[lazywking]

not same

[lazywking]

if I create one Agent application(simple way, not using agent node in workflow chat type), call same tools, which worked fine. the problem should in agent node scripts.
2025-04-09 09:06:45.140 DEBUG [Thread-1197 (_generate_worker)] [connectionpool.py:544] - http://plugin_daemon:5002 "POST /plugin/9e3c5683-bc96-4e2d-b67f-34be8a56929f/dispatch/llm/num_tokens HTTP/1.1" 200 None

[on_llm_before_invoke]
Model: deepseek-r1
Parameters:
stop: ['Observation']
Stream: True
User: b72e1db4-083a-471c-a257-f7d3a8649230
Prompt messages:
role: system
content: Respond to the human as helpfully and accurately as possible.

你的任务是处理账号相关的问题，其他问题不需要处理

You have access to the following tools:

[{"name": "checkUserAccount", "description": "\u68c0\u67e5\u63d0\u4f9b\u7684\u7528\u6237\u540d\u662f\u5426\u5b58\u5728\u6216\u6709\u6548 (Check if the provided username exists or is valid)", "parameters": {"type": "object", "properties": {"username": {"type": "string", "description": "\u9700\u8981\u68c0\u67e5\u7684\u7528\u6237\u540d (The username to check)"}}, "required": ["username"]}}]

Use a json blob to specify a tool by providing an action key (tool name) and an action_input key (tool input).
Valid "action" values: "Final Answer" or checkUserAccount

Provide only ONE action per $JSON_BLOB, as shown:

{
  "action": $TOOL_NAME,
  "action_input": $ACTION_INPUT
}

Follow this format:

Question: input question to answer
Thought: consider previous and subsequent steps
Action:

$JSON_BLOB

Observation: action result
... (repeat Thought/Action/Observation N times)
Thought: I know what to respond
Action:

{
  "action": "Final Answer",
  "action_input": "Final response to human"
}

Begin! Reminder to ALWAYS respond with a valid json blob of a single action. Use tools if necessary. Respond directly if appropriate. Format is Action:$JSON_BLOBthen Observation:.

role: user
content: 我的用户名是wa请查看账号是否正常