← Back to Skills Marketplace
confidentkai

Tavily Search Optimized

by kvs-GoN · GitHub ↗ · v1.0.3 · MIT-0
cross-platform ⚠ suspicious
82
Downloads
0
Stars
0
Active Installs
4
Versions
Install in OpenClaw
/install tavily-search-optimized-v1
Description
Web search via Tavily API (alternative to Brave). Use when the user asks to search the web / look up sources / find links and Brave web_search is unavailable...
Usage Guidance
This skill implements a Tavily web-search client and appears to be mostly coherent, but check these points before installing: - Provide an API key: the scripts require TAVILY_API_KEY (or TAVILY_KEY). The registry metadata did not declare this requirement — set the env var or add it to ~/.openclaw/.env before use. - Review where secrets are stored: examples suggest writing your API key to ~/.openclaw/.env in plaintext. Ensure that file has safe permissions and that you are comfortable storing the key there. - Confirm the Tavily endpoint: the code posts to https://api.tavily.com. If you don't trust that service, do not provide credentials. - Be aware of local caching: search results are cached under ~/.openclaw/cache/tavily (this writes files to your home directory). If that is a concern, run with --no-cache or set a custom cache dir. - Minor docs/code mismatches: examples reference formats or paths (e.g., /root/.openclaw/...) that may not match your environment — inspect and possibly edit the script paths before running. If you want higher assurance, review the full scripts (they are included) for any further unexpected network calls or data-handling behavior, or run them in an isolated environment first.
Capability Analysis
Type: OpenClaw Skill Name: tavily-search-optimized-v1 Version: 1.0.3 The skill bundle provides a well-documented and functional tool for performing web searches via the Tavily API. The Python scripts (tavily_search.py and tavily_search_legacy.py) use standard libraries for network requests and implement a local file-based caching system in a dedicated directory (~/.openclaw/cache/tavily). There is no evidence of data exfiltration, malicious execution, or prompt injection; the code only accesses its own configuration file (~/.openclaw/.env) and the official Tavily API endpoint (api.tavily.com).
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
Name/description claim a web-search client for Tavily, and the code implements that. However the registry metadata lists no required environment variables or primary credential, while the SKILL.md and both scripts require a TAVILY_API_KEY (or alias TAVILY_KEY). That mismatch (metadata says none required but the code will fail without an API key) is an incoherence that should be resolved.
Instruction Scope
Runtime instructions and examples are limited to performing searches, formatting results, and caching. The SKILL.md and scripts read environment variables and ~/.openclaw/.env, create/read cache files under ~/.openclaw/cache/tavily, and make POST requests to https://api.tavily.com. There is no evidence of attempts to read unrelated system files or exfiltrate arbitrary secrets, but examples use absolute /root/.openclaw paths (which may be incorrect for non-root deployments) and an example shows using subprocess to call the script from other code — both are benign but worth noticing.
Install Mechanism
This is instruction + bundled scripts only; there is no install spec, no external downloads, and no packaged post-install code. No high-risk install behavior observed.
Credentials
The code legitimately needs a Tavily API key (TAVILY_API_KEY or TAVILY_KEY) and optional cache/config env vars (TAVILY_CACHE_DIR, TAVILY_DEFAULT_TIMEOUT, etc.). Those are proportionate to a search client. The concern is that the registry metadata did not declare the API key requirement, which could cause accidental failure or secrecy mistakes. SKILL.md and scripts also recommend adding the API key to ~/.openclaw/.env or exporting it — storing secrets in a plaintext file is common but should be done with caution (file permissions).
Persistence & Privilege
The skill does not request always:true and will not be force-enabled; it writes cache files under ~/.openclaw/cache/tavily and may create ~/.openclaw/.env if the user follows examples, which is within the skill's scope and not a system-wide privilege escalation. No modifications to other skills or global agent settings observed.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install tavily-search-optimized-v1
  3. After installation, invoke the skill by name or use /tavily-search-optimized-v1
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.3
重要修复:修复配置文件读取逻辑,现在正确读取~/.openclaw/.env中的配置,实现环境变量>配置文件>默认值的正确优先级
v1.0.2
修复类型错误:修复参数为None时的类型错误,增强参数处理逻辑,添加详细模式参数显示
v1.0.1
修复配置不一致问题:现在所有config.example.env中的环境变量都在脚本中实际使用,包括TAVILY_CACHE_DIR、TAVILY_DEFAULT_TIMEOUT、TAVILY_CACHE_TTL、TAVILY_MAX_RESULTS、TAVILY_SEARCH_DEPTH
v1.0.0
优化版本:添加缓存系统、增强错误处理、多种输出格式、配置选项、完整文档和示例
Metadata
Slug tavily-search-optimized-v1
Version 1.0.3
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 4
Frequently Asked Questions

What is Tavily Search Optimized?

Web search via Tavily API (alternative to Brave). Use when the user asks to search the web / look up sources / find links and Brave web_search is unavailable... It is an AI Agent Skill for Claude Code / OpenClaw, with 82 downloads so far.

How do I install Tavily Search Optimized?

Run "/install tavily-search-optimized-v1" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Tavily Search Optimized free?

Yes, Tavily Search Optimized is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Tavily Search Optimized support?

Tavily Search Optimized is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Tavily Search Optimized?

It is built and maintained by kvs-GoN (@confidentkai); the current version is v1.0.3.

💬 Comments