Skip to content

Align Tavily Search schema with current API behavior #2720

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
yasu-oh wants to merge 6 commits into
base: main
Choose a base branch
from
Open

Align Tavily Search schema with current API behavior #2720

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
462fb1a
feat: update Tavily search tool schema to align with current API
yasu-oh Mar 12, 2026
2f6c907
fix: align chunks_per_source behavior with Tavily search_depth options
yasu-oh Mar 12, 2026
7a7c0fb
Merge branch 'main' into main
crazywoola Mar 13, 2026
5d0b477
Merge branch 'main' into main
yasu-oh Mar 13, 2026
4b39b77
Merge branch 'main' into main
yasu-oh Mar 26, 2026
46468e6
Merge branch 'main' into main
yasu-oh Mar 31, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion models/gemini/.difyignore
View file
Open in desktop

This file was deleted.

1 change: 1 addition & 0 deletions models/gemini/.difyignore
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
../../.difyignore
2 changes: 1 addition & 1 deletion tools/tavily/pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
[project]
name = "tavily"
version = "0.1.3"
version = "0.1.5"
description = "A powerful AI-native search engine and web content extraction tool for Dify"
readme = "README.md"
requires-python = ">=3.12"
Expand Down
5 changes: 4 additions & 1 deletion tools/tavily/tools/tavily_search.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,8 +108,11 @@ def _process_params(self, params: dict[str, Any]) -> dict:
processed_params.setdefault("search_depth", "basic")
processed_params.setdefault("topic", "general")
processed_params.setdefault("max_results", 5)
if processed_params.get("search_depth") == "advanced":

if processed_params.get("search_depth") in {"advanced", "fast"}:
processed_params.setdefault("chunks_per_source", 3)
else:
processed_params.pop("chunks_per_source", None)
return processed_params


Expand Down
23 changes: 11 additions & 12 deletions tools/tavily/tools/tavily_search.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ description:
real-time, accurate, and factual results at speed.
zh_Hans: 专为人工智能代理 (LLM) 构建的搜索引擎工具，可快速提供实时、准确和真实的结果。
llm:
A tool for search engine built specifically for AI agents (LLMs), delivering
A search engine tool built specifically for AI agents (LLMs), delivering
real-time, accurate, and factual results at speed.
extra:
python:
Expand Down Expand Up @@ -66,7 +66,7 @@ parameters:
label:
en_US: Topic
zh_Hans: 主题
llm_description: The category of the search. Options include 'general', 'news', or 'finance'.
llm_description: The category of the search. Valid options are 'general', 'news', and 'finance'.
options:
- label:
en_US: General
Expand All @@ -90,7 +90,7 @@ parameters:
label:
en_US: Time Range
zh_Hans: 时间范围
llm_description: The time range back from the current date to filter results. Options include 'not_specified', 'day', 'week', 'month', or 'year'.
llm_description: The time range back from the current date to filter results. Valid options are 'not_specified', 'day', 'week', 'month', and 'year'. When time_range is set, do not set start_date or end_date.
options:
- label:
en_US: Not Specified
Expand Down Expand Up @@ -123,7 +123,7 @@ parameters:
label:
en_US: Start Date
zh_Hans: 开始日期
llm_description: The start date for filtering results. Required to be written in the format YYYY-MM-DD. Results will be filtered to include content from this date onwards.
llm_description: The start date for filtering results. Required to be written in the format YYYY-MM-DD. Results will be filtered to include content from this date onwards. When time_range is set, do not set start_date or end_date.
required: false
type: string
- name: end_date
Expand All @@ -134,7 +134,7 @@ parameters:
label:
en_US: End Date
zh_Hans: 结束日期
llm_description: The end date for filtering results. Required to be written in the format YYYY-MM-DD. Results will be filtered to include content up to this date.
llm_description: The end date for filtering results. Required to be written in the format YYYY-MM-DD. Results will be filtered to include content up to this date. When time_range is set, do not set start_date or end_date.
required: false
type: string
- name: auto_parameters
Expand Down Expand Up @@ -194,7 +194,7 @@ parameters:
label:
en_US: Include Answer
zh_Hans: 包含答案
llm_description: When set to true or 'basic', includes a short answer to the original query in the response, 'advanced' for a more detailed comprehensive answer.
llm_description: When set to true or 'basic', includes a short answer to the original query in the response. Use 'advanced' for a more detailed, comprehensive answer.
options:
- label:
en_US: "False"
Expand Down Expand Up @@ -273,25 +273,25 @@ parameters:
default: 3
form: form
human_description:
en_US: Maximum number of content chunks returned per source (advanced search only).
zh_Hans: 每个来源返回的内容片段的最大数量（仅限高级搜索）。
en_US: Maximum number of content chunks returned per source (advanced and fast search only).
zh_Hans: 每个来源返回的内容片段的最大数量（仅限高级和快速搜索）。
label:
en_US: Chunks Per Source
zh_Hans: 每来源片段数
llm_description: Maximum number of relevant content chunks (1-3) returned per source. Valid only when search_depth is advanced.
llm_description: Maximum number of relevant content chunks (1-3) returned per source. Valid when search_depth is 'advanced' or 'fast'.
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist bot Mar 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

This description update correctly reflects that chunks_per_source is valid for 'fast' search. However, the backing implementation in tools/tavily/tools/tavily_search.py appears to be inconsistent with this change. The _process_params method only sets a default value for chunks_per_source when search_depth is 'advanced':

# tools/tavily/tools/tavily_search.py:111-112
if processed_params.get("search_depth") == "advanced":
    processed_params.setdefault("chunks_per_source", 3)

This logic does not include the 'fast' search depth. As a result, the default value for chunks_per_source will not be applied for 'fast' searches, which may lead to behavior that is inconsistent with the updated description. Please consider aligning the Python code with this schema change.

max: 3
min: 1
required: false
type: number
- name: country
form: llm
human_description:
en_US: Boost search results from a specific country (topic must be general).
en_US: Boost search results for a specific country (topic must be general).
zh_Hans: 从特定国家/地区提升搜索结果（主题必须是general）。
label:
en_US: Country
zh_Hans: 国家
llm_description: Select a country to boost its content in results (only when topic is general).
llm_description: Select a country to boost results only when topic is general. Use one of the exact option values.
required: false
type: select
options:
Expand Down Expand Up @@ -959,7 +959,6 @@ parameters:
en_US: Zimbabwe
zh_Hans: 津巴布韦
value: zimbabwe
default: united states
- name: include_domains
form: llm
human_description:
Expand Down
Loading