Liang Tavily Search 1.0.1

Web search using Tavily's LLM-optimized API. Returns relevant results with content snippets, scores, and metadata.

Security Scan

The skill's code, instructions, and required credential (TAVILY_API_KEY) match its stated purpose (calling Tavily's search API); nothing in the files indicates unrelated or hidden capabilities.

Purpose & Capability

Name/description, required binary (node), and required env var (TAVILY_API_KEY) align with a simple Node script that calls https://api.tavily.com/search. The script only talks to the Tavily API and formats results for display.

Instruction Scope

SKILL.md and the script are narrowly scoped to running the search script and using the Tavily API. Note: the script supports a --raw-content option which requests full page content from Tavily; that increases the amount of user data sent to the external API when used. Otherwise the instructions do not read unrelated files or other env vars.

Install Mechanism

No install spec — instruction-only plus a small included script. No downloads or archives; nothing is written to disk by an installer. The requirement that 'node' be present is appropriate for the .mjs script.

Credentials

Only TAVILY_API_KEY is required and declared as the primary credential. That is proportional for a service-backed search skill. No other secrets, config paths, or unrelated credentials are requested.

Persistence & Privilege

The skill does not request always:true and does not modify agent/system configs. It can be invoked by the agent (normal default), but there are no other privileged behaviors or persistent privileges requested.

Assessment

This skill appears coherent and limited to calling Tavily's search API. Before installing: 1) Be comfortable trusting Tavily with any full-page content you request (avoid --raw-content if you don't want large page contents sent to the service). 2) Store TAVILY_API_KEY securely (env var or the platform's secret store); the script uses a bearer token to call the API. 3) Ensure your runtime has a recent Node version that supports fetch (the script uses top-level await and fetch). 4) Note a minor metadata mismatch in owner IDs between registry metadata and _meta.json (likely benign but FYI). If you need stricter guarantees about data handling, review Tavily's privacy policy or avoid enabling --raw-content.

Tavily Search

Search the web and get relevant results optimized for LLM consumption.

Authentication

Get your API key at https://tavily.com and add to your OpenClaw config:

{
  "skills": {
    "entries": {
      "tavily-search": {
        "enabled": true,
        "apiKey": "tvly-YOUR_API_KEY_HERE"
      }
    }
  }
}

Or set the environment variable:

export TAVILY_API_KEY="tvly-YOUR_API_KEY_HERE"

Quick Start

Using the Script

node {baseDir}/scripts/search.mjs "query"
node {baseDir}/scripts/search.mjs "query" -n 10
node {baseDir}/scripts/search.mjs "query" --deep
node {baseDir}/scripts/search.mjs "query" --topic news

Examples

# Basic search
node {baseDir}/scripts/search.mjs "python async patterns"

# With more results
node {baseDir}/scripts/search.mjs "React hooks tutorial" -n 10

# Advanced search
node {baseDir}/scripts/search.mjs "machine learning" --deep

# News search
node {baseDir}/scripts/search.mjs "AI news" --topic news

# Domain-filtered search
node {baseDir}/scripts/search.mjs "Python docs" --include-domains docs.python.org

Options

Option	Description	Default
-n <count>	Number of results (1-20)	10
--depth <mode>	Search depth: ultra-fast, fast, basic, advanced	basic
--topic <topic>	Topic: general or news	general
--time-range <range>	Time range: day, week, month, year	-
--include-domains <domains>	Comma-separated domains to include	-
--exclude-domains <domains>	Comma-separated domains to exclude	-
--raw-content	Include full page content	false
--json	Output raw JSON	false

Search Depth

Depth	Latency	Relevance	Use Case
ultra-fast	Lowest	Lower	Real-time chat, autocomplete
fast	Low	Good	Need chunks but latency matters
basic	Medium	High	General-purpose, balanced
advanced	Higher	Highest	Precision matters, research

Tips

• Keep queries under 400 characters - Think search query, not prompt
• Break complex queries into sub-queries - Better results than one massive query
• Use --include-domains to focus on trusted sources
• Use --time-range for recent information
• Filter by score (0-1) to get highest relevance results