Replies: 6 comments 4 replies
-
|
To develop and deploy an agent strategy plugin in Dify, running it with Dify does not currently support true hot reloading for agent strategy plugins—meaning, if you change your plugin code, you need to manually restart the plugin process for changes to take effect. You do not need to restart the whole Dify platform, just the plugin process itself [source]. For production, package your plugin (e.g., with Official docs: If you run into issues, double-check your To reply, just mention @dosu. How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other
|
Beta Was this translation helpful? Give feedback.
-
|
python -m main {"event": "log", "data": {"level": "INFO", "message": "Installed agent: agent", "timestamp": 1767840496.2360218}} 我在本地启动了python -m main，看到上述消息，但是在dify的插件页面，我没看到这个安装后的插件，是什么配置错误了吗？ |
Beta Was this translation helpful? Give feedback.
-
|
你的插件没有显示在 Dify 插件页面，是因为仅仅用
如果还是不显示，建议开启 LOG_LEVEL=DEBUG，检查插件 daemon 和 Dify 后台日志，确认插件注册和 schema 校验是否有报错 [参考]。 简单来说， To reply, just mention @dosu. How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other
|
Beta Was this translation helpful? Give feedback.
-
|
谢谢，我解决了这个问题，但是我的agent strategy 部署上去后，执行后，我在docker-api-1中发现显示如下的错误，这个大致是什么原因？ ############################################################################# |
Beta Was this translation helpful? Give feedback.
-
|
这个错误
排查建议：
这个错误不是 agent strategy 代码本身导致的，而是插件运行环境或注册流程有问题。升级 Dify 和 plugin daemon 到最新版本，并严格按照官方流程安装插件，可以避免类似问题 [参考5]。 To reply, just mention @dosu. How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other
|
Beta Was this translation helpful? Give feedback.
-
|
但是我已经添加上了，在插件和chatflow中我都已经可以选择了。我在agent节点中可以选择我添加的agent strategy了。那不是应该节点已经添加成功了吗？现在说的这个是指什么没找到？ 我按照python -m main方式安装后，在命令行窗口没有显示错误 
|
Beta Was this translation helpful? Give feedback.
-
|
你能在 UI 和 chatflow 里选择 agent strategy，只说明插件的“注册信息”和 schema 已经被 Dify 后台识别，但这并不代表插件的运行环境（runtime）真的可用。Dify 的插件架构分为“注册/展示”（UI 可选）和“运行时调度”两部分，只有当 plugin daemon 能找到并启动对应的插件 runtime 节点时，插件功能才能真正被调用。 你遇到的 “no available node, plugin runtime not found” 错误，意思是 plugin daemon 没有找到可用的插件运行节点来执行你的 agent strategy。常见原因包括：
建议你按以下顺序排查：
这个问题不是 agent strategy 代码本身导致的，而是 plugin daemon 没有找到可用的插件 runtime 节点来执行请求。只有 plugin daemon 端的 runtime 节点健康、插件包完整、配置无误，插件功能才能正常运行。详细原理和排查建议可参考官方代码实现和社区经验和插件 daemon 错误处理逻辑。 To reply, just mention @dosu. How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other
|
Beta Was this translation helpful? Give feedback.
-
|
当我执行python -m main的时候，我在docker-plugin_daemon-1中看到如下的信息，这是否可以确认已经正确安装了 2026/01/08 03:54:33 /app/internal/db/executor.go:259 record not found 2026/01/08 03:54:33 /app/internal/db/executor.go:259 record not found |
Beta Was this translation helpful? Give feedback.
-
|
我在docker api -1的 log中看到如下很多的错误日志，这是说明了哪个方面的错误导致的 2026-01-08 11:57:20.122 ERROR [GraphWorker-1] [base.py:277] - Error in stream response for plugin {'code': -500, 'message': '{"message":"{\"args\":{},\"error_type\":\"ValidationError\",\"message\":\"8 validation errors for AgentInvokeMessage\\ntype\\n Field required [type=missing, input_value={'message': 'Error: 1 val...tic.dev/2.12/v/missing'}, input_type=dict]\\n For further information visit https://errors.pydantic.dev/2.12/v/missing\\\\nmessage.TextMessage\\\\n Input should be a valid dictionary or instance of TextMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\\\nmessage.JsonMessage\\\\n Input should be a valid dictionary or instance of JsonMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\\\nmessage.VariableMessage\\\\n Input should be a valid dictionary or instance of VariableMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\\\nmessage.BlobMessage\\\\n Input should be a valid dictionary or instance of BlobMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\\\nmessage.BlobChunkMessage\\\\n Input should be a valid dictionary or instance of BlobChunkMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\\\nmessage.LogMessage\\\\n Input should be a valid dictionary or instance of LogMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\\\nmessage.RetrieverResourceMessage\\\\n Input should be a valid dictionary or instance of RetrieverResourceMessage [type=model_type, input_value='Error: 1 validation erro...ntic.dev/2.12/v/missing', input_type=str]\\n For further information visit https://errors.pydantic.dev/2.12/v/model_type\\"}","error_type":"PluginInvokeError","args":null}', 'data': None} ** |
Beta Was this translation helpful? Give feedback.
-
|
你的 agent strategy 插件报错 “ValidationError for AgentInvokeMessage”，说明插件返回的数据结构不符合 Dify 要求的 schema。Dify 期望插件返回的数据必须是一个包含 type 和 message 字段的结构化对象，且 message 必须是指定类型的字典（如 TextMessage、JsonMessage、VariableMessage 等），不能直接返回字符串或未按要求组织的数据。 举例，正确的返回格式如下：
常见错误是直接返回字符串、缺少 type 字段、message 字段不是字典或缺少必填项。你需要确保插件的 yield/return 语句严格按照 schema 组织，不能直接 yield "some string"，而是 yield { "type": ..., "message": {...} }。 详细 schema 和更多例子见官方文档：Agent strategy plugin开发规范 建议：
只要返回结构正确，ValidationError 就会消失，插件即可正常工作。 To reply, just mention @dosu. How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other
|
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Self Checks
1. Is this request related to a challenge you're experiencing? Tell me about your story.
I want to develop a private agetn strategy. I want to hot to deploy it like a normal plugin with "python -m main". Is this correct?
2. Additional context or comments
No response
Beta Was this translation helpful? Give feedback.
All reactions