Custom tool plugin parameter passing error

[libin1219]

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

Here are the contents of the video-deal.yaml file for my tool plugin
identity:
name: "video-deal"
author: "libin"
label:
en_US: "video-deal"
zh_Hans: "视频切割"
pt_BR: "video-deal"
ja_JP: "video-deal"
description:
human:
en_US: "Cut video into segments by specified duration (via video file URL)"
zh_Hans: "按指定时长切割视频（通过视频文件URL）"
pt_BR: "Video segmentation and extraction of video sound"
ja_JP: "Video segmentation and extraction of video sound"
llm: "Cut video into segments by specified duration via URL"
parameters:

• name: video_source
  type: string
  required: true
  label:
  en_US: video_source
  zh_Hans: 视频文件URL
  pt_BR: video_source
  human_description:
  en_US: "Video file access URL (from uploaded file)"
  zh_Hans: "视频文件访问URL（用户上传视频的临时链接"
  pt_BR: "Video segmentation and extraction of video sound"
  ja_JP: "Video segmentation and extraction of video sound"
  llm_description: "Video file access URL (uploaded by user)"
  form: form
• name: segment_duration
  type: number
  required: false
  default: 300
  label:
  zh_Hans: 分段时长
  en_US: Segment Duration
  human_description:
  zh_Hans: "视频分段的时长（单位：秒），默认300秒（5分钟），最小值10秒，最大值3600秒（1小时）"
  en_US: "Video segment duration (seconds), default 300s (5min), min 10s, max 3600s (1h)"
  llm_description: "Video segment duration in seconds, min 10, max 3600, default 300"
  form: form
  placeholder:
  zh_Hans: "请输入分段时长（默认300）"
  validation:
  min: 10
  max: 3600
  extra:
  python:
  source: tools/video-deal.py

[dosubot[bot]]

There are a few common causes for parameter passing errors in Dify custom tool plugins:

• The parameter names and types in your YAML must exactly match what your Python implementation expects. If there’s any mismatch, or if a required parameter is missing, Dify will raise an error like tool parameter ... not found in tool config during initialization or runtime. Double-check that your Python code expects video_source (string, required) and segment_duration (number, optional, default 300) as defined in your YAML, and that it handles defaults and validation constraints properly example.

• If you’re running Dify self-hosted via Docker and using versions 1.3.7 to 1.9.1, there’s a known bug where custom tool parameters may not be passed correctly at runtime—parameters set in tool settings can be ignored and missing in the request payload details. If this matches your setup, upgrading to a newer version or checking for related fixes may help.

• Only primitive types like string and number are supported for custom tool input fields in the UI and YAML. Array/object types are not supported and will cause errors reference.

• Make sure your YAML is valid and follows the structure of official plugins. Formatting errors or missing required fields can cause both backend and frontend errors.

If you can share the exact error message or describe what happens when the error occurs, and confirm how your Python implementation receives parameters, I can help pinpoint the issue further. Also, reviewing backend logs and browser console errors can provide clues if the UI fails to render or parameters are missing.

To reply, just mention @dosu.

---

Docs are dead. Just use Dosu.

[libin1219]

I am using version 1.10.1 of dify. Does it require an upgrade to solve this problem

[dosubot[bot]]

Yes, upgrading Dify is recommended to resolve custom tool plugin parameter passing issues. Version 1.10.1 has known bugs with optional parameters and manual/default value handling that can cause validation errors or missing parameters at runtime. These issues are fixed in later versions—upgrading will ensure optional parameters work as expected and manual/default values are correctly passed to your plugin (source, source, source).

After upgrading, republish your workflow nodes and refresh your browser to make sure the changes take effect. If you run into migration errors or plugin failures after upgrading, let me know—there are some manual steps for database migration in self-hosted setups (details).

To reply, just mention @dosu.

---

Share context across your team and agents. Try Dosu.

[libin1219]

I upgraded dify to 1.11.4, but still got an error

[libin1219]

[dosubot[bot]]

The “invalid variable” error in Dify 1.11.x usually means there’s a mismatch or missing field in your tool’s parameter or output schema. In recent versions, Dify enforces stricter type validation: string/select parameters must receive a string, and every workflow output must include a value_type field (like "string", "number", "object")—missing this causes validation errors in the UI and when publishing tools (source) (source).

Here’s what to check:

1. Make sure your YAML is strictly formatted—no extra fields, correct indentation, and all required fields present.
2. If your workflow was created in an older Dify version, manually add value_type to each output variable (for example, value_type: "string") in your workflow JSON or database, or re-import the end node to auto-populate missing fields.
3. Republish your tool node and refresh/clear your browser cache to ensure the latest schema is loaded.
4. If you still see the error, check backend logs and browser console for more details and share the exact error message.

If you’re using Hengine, try testing the plugin in a plain Dify instance to rule out integration-specific issues. If the problem persists, share a minimal YAML/plugin example and error logs for deeper analysis.

To reply, just mention @dosu.

---

Docs are dead. Just use Dosu.