← Back to Skills Marketplace
moochmaniac

Tavily Web Search

by Patrick · GitHub ↗ · v1.0.0
cross-platform ⚠ suspicious
3514
Downloads
6
Stars
4
Active Installs
1
Versions
Install in OpenClaw
/install moochmaniac-tavily-search
Description
Real-time web search using Tavily API optimized for AI agents. Use when you need current information, recent news, factual research, competitive analysis, or...
Usage Guidance
This skill is coherent with its stated purpose (Tavily web search) and not obviously malicious, but there are implementation issues you should address before trusting it: 1) Provide your Tavily API key (either set TAVILY_API_KEY or place the key at the path you control). Note the script looks for several paths and one path in the code ("~/.openclaw/workspace/workspace/secrets/...") does not match the SKILL.md — confirm which location will actually be used. 2) The script contains a bug: if no secret file is found, it will crash instead of falling back to the environment variable. Fix the os.path.exists(secrets_path) logic to guard when secrets_path is None. 3) Ensure the Python 'requests' package is installed in the runtime environment (the skill does not declare this dependency). 4) Verify you trust the Tavily service and its endpoint (https://api.tavily.com/search) because search queries and API key will be sent there. If you are not comfortable exposing queries or a key to that endpoint, do not install. If you plan to use it, either fix the code or run it in an isolated environment and supply the API key via a location you control.
Capability Analysis
Type: OpenClaw Skill Name: moochmaniac-tavily-search Version: 1.0.0 The skill is a standard implementation of a Tavily Search API wrapper for AI agents. It correctly handles API key retrieval from local secret files or environment variables and communicates only with the official Tavily endpoint (api.tavily.com). While the script in `scripts/search.py` contains a minor logic bug (a TypeError that occurs if no secret file is found, preventing the environment variable fallback), there is no evidence of malicious intent, data exfiltration, or prompt injection. The tool's behavior is entirely consistent with its stated purpose of providing real-time web search capabilities.
Capability Assessment
Purpose & Capability
The name/description match the included script: it calls a Tavily search endpoint and returns structured results. Requiring a Tavily API key is proportionate to the stated purpose. However the registry metadata lists no required env vars while the script expects an optional TAVILY_API_KEY env var or workspace secret — a minor inconsistency. No unrelated credentials or capabilities are requested.
Instruction Scope
The SKILL.md and script instruct the agent to read an API key from workspace/secrets or a relative secrets path or an environment variable — that is expected for an API-backed search. However the script has a logic bug: if none of the candidate secret files exist, secrets_path remains None and the subsequent os.path.exists(secrets_path) will raise an exception (TypeError), preventing the intended fallback to the TAVILY_API_KEY environment variable. Also the script references an odd duplicated path (~/.openclaw/workspace/workspace/...) that does not match the SKILL.md path text, which could cause confusion about where keys should be stored.
Install Mechanism
This is an instruction-only skill with a bundled Python script (no install spec). The script imports the third-party 'requests' library but the skill does not declare dependency installation instructions, which may lead to runtime failures if requests isn't available. No remote downloads or high-risk install behaviors are present.
Credentials
The only secret material the skill needs is a Tavily API key (file or TAVILY_API_KEY env var), which is proportional to a web search integration. The registry metadata claiming 'no required env vars' is inconsistent with the script/README which document TAVILY_API_KEY and workspace secret paths. The secret paths referenced are limited to workspace/relative locations rather than unrelated system credentials.
Persistence & Privilege
The skill does not request always:true, does not modify other skills or system-wide settings, and has normal user-invocable/autonomous invocation defaults. No elevated persistence or cross-skill configuration changes are present.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install moochmaniac-tavily-search
  3. After installation, invoke the skill by name or use /moochmaniac-tavily-search
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release: Real-time web search via Tavily API with structured results, relevance scores, and AI-generated answers
Metadata
Slug moochmaniac-tavily-search
Version 1.0.0
License
All-time Installs 4
Active Installs 4
Total Versions 1
Frequently Asked Questions

What is Tavily Web Search?

Real-time web search using Tavily API optimized for AI agents. Use when you need current information, recent news, factual research, competitive analysis, or... It is an AI Agent Skill for Claude Code / OpenClaw, with 3514 downloads so far.

How do I install Tavily Web Search?

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

Is Tavily Web Search free?

Yes, Tavily Web Search is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Tavily Web Search support?

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

Who created Tavily Web Search?

It is built and maintained by Patrick (@moochmaniac); the current version is v1.0.0.

💬 Comments