Agent skill

Tavily

by Mathews-Tom in Mathews-Tom/armory

AI-optimized web search via Tavily API when no URL is provided.

MITAuto-check passedProductivity & Automation

Install Tavily

skills CLI
$ npx skills add Mathews-Tom/armory --skill tavily -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install Mathews-Tom/armory tavily --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/Mathews-Tom/armory.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/tavily .claude/skills/tavily && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
tavily
GitHub stars
328
Token cost
~1.5k tokens
SKILL.md length
500 words
Files
4 (incl. scripts)
Skills in repo
80
Repo updated
First seen
Licence
MIT

At a glance

AI-optimized web search via Tavily API when no URL is provided.

  • : search the web
  • SKILL.md covers When to Use Tavily vs Other…, Scripts, Search and Extract Content from URL, plus 3 more sections
  • Runs JavaScript scripts from its folder; calls node; reaches source.com; needs TAVILY_API_KEY
  • Find recent information

What it does

Tavily is an agent skill from Mathews-Tom/armory. AI-optimized web search via Tavily API when no URL is provided. Returns clean AI-ready snippets for current info and news. Triggers on: "search the web", "look up online", "find recent information", "web search". NOT for known URLs, use web-fetch.

Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 5 other files, including scripts (for example `evals/cases.yaml`).

It sits in Productivity & Automation, covering Web search. It works with Tavily. The repository describes itself as: Curated, production-grade skills for AI coding agents. Battle-tested workflows for developers who use AI seriously. The licence is MIT.

When your agent uses it

  • : search the web
  • Find recent information

Example prompts

  • “search the web”
  • “look up online”
  • “find recent information”
  • “/tavily”

Requirements

  • Node.js
  • A credential in TAVILY_API_KEY

What it can do on your machine

Read from SKILL.md and the folder at commit 4594fb7. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 2 files in scripts/ (JavaScript), which the agent can run.

    Shell commands in SKILL.md call:

    • node

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • source.com

    Also links to:

    • tavily.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names these keys or tokens, usually read from environment variables:

    • TAVILY_API_KEY

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Tavily loads about 1.5k tokens when it runs. Until then it costs about 64 tokens; SKILL.md has 500 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~64
When it runs · the whole SKILL.md, loaded when a task matches
~1.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from Mathews-Tom/armory at commit 4594fb7, republished under its MIT licence (© Mathews-Tom). 500 words, ~1,504 tokens.

Download SKILL.mdSave it as .claude/skills/tavily/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.
name
tavily
description
AI-optimized web search via Tavily API when no URL is provided. Returns clean AI-ready snippets for current info and news. Triggers on: "search the web", "look up online", "find recent information", "web search". NOT for known URLs, use web-fetch.
metadata.version
1.1.1
metadata.category
research
metadata.tags
web-search, research, real-time, ai-optimized
metadata.difficulty
beginner

Web search and URL content extraction powered by Tavily API. Returns clean, structured results optimized for AI consumption — no HTML noise, no cookie banners, no pagination.

When to Use Tavily vs Other Tools

SituationToolReason
Need current information on a topicsearch.mjsTavily aggregates and ranks live results
Need news from the last N dayssearch.mjs --topic news --days NNews mode filters recency
Have a specific URL, need full textextract.mjsExtracts clean article content from URL
Need documentation for a librarysearch.mjs --deepDeep mode covers docs sites thoroughly
Need a static resource already knowndirect fetch or readSkip Tavily if no live lookup needed
Need multiple competing perspectivessearch.mjs -n 10Increase result count for breadth

Scripts

Both scripts exist at scripts/ and require TAVILY_API_KEY in the environment.

  • scripts/search.mjs — query-based web search
  • scripts/extract.mjs — full content extraction from a URL
bash
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
node {baseDir}/scripts/search.mjs "query" --topic news --days 3
Search Options
OptionDefaultDescription
-n <count>5Number of results to return (max: 20). Increase for broader coverage.
--deepoffAdvanced search mode — slower but more comprehensive. Use for complex research questions, technical deep-dives, or when standard results are insufficient.
--topic <topic>generalgeneral for standard search; news for current events filtered by recency.
--days <n>—Only valid with --topic news. Limits results to articles published in the last n days.

Extract Content from URL

bash
node {baseDir}/scripts/extract.mjs "https://example.com/article"

Use when you already have a URL and need the full page content — Tavily strips navigation, ads, and boilerplate, returning structured article text.

Output Format

Search results return a JSON array. Each result contains:

json
[
  {
    "title": "Article or page title",
    "url": "https://source.com/path",
    "content": "Clean text snippet (AI-optimized excerpt)",
    "score": 0.92
  }
]

score is Tavily's relevance ranking (0–1). Higher is more relevant. When parsing results, use content for summarization and url for attribution. Do not treat content as the full article — use extract.mjs on the URL if complete text is needed.

Extract output returns a single object:

json
{
  "url": "https://source.com/path",
  "raw_content": "Full extracted page text..."
}
Show full SKILL.md (198 more words)Show less

Error Handling

ErrorCauseResolution
TAVILY_API_KEY not setEnvironment variable missingSet TAVILY_API_KEY before running. Get a key at https://tavily.com
401 UnauthorizedInvalid API keyVerify the key value; regenerate at Tavily dashboard if needed
429 Too Many RequestsRate limit hitReduce request frequency; free tier allows 1000 API credits/month
Empty results array []No matches foundBroaden or rephrase the query; try --deep for niche topics
ETIMEDOUT / timeoutNetwork or API latencyRetry once; --deep searches take longer (up to 30s)
Extract returns empty contentPage blocks scrapersUse a different source or fall back to a manual fetch approach

Limitations

  • Rate limits: Free tier is 1000 API credits/month. Each search costs 1 credit; --deep costs more.
  • Content freshness: Results reflect Tavily's index; very recent content (last few hours) may not appear in general mode. Use --topic news for recency-sensitive queries.
  • Result quality: Tavily optimizes for AI-readable content but cannot guarantee factual accuracy. Verify critical claims from primary sources.
  • Paywalled content: Articles behind paywalls return only the publicly accessible excerpt.
  • Non-text content: PDFs, images, and media pages extract poorly or not at all.
  • Language: Best results for English queries. Other languages are supported but result quality varies.

© Mathews-Tom, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (scripts) in skills/tavily of Mathews-Tom/armory.

  • SKILL.md
  • evals/cases.yaml
  • scripts/extract.mjs
  • scripts/search.mjs

Open the folder on GitHubat commit 4594fb7

Compare with similar skills

Tavily next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Tavily compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Tavily this skillMathews-Tom/armory328—~1.5kAutomated safety check: PassMIT
Ultimate Searchckckck/UltimateSearchSkill289—~944Automated safety check: NotesMIT
Web SearchEXboys/skilllite1702 repos~1kAutomated safety check: PassMIT
Mysearchskernelx/MySearch-Proxy159—~3kAutomated safety check: NotesNone
Tavilyopenclaw/openclaw392k2 repos~1.2kAutomated safety check: PassMIT
Web Search Plus Plugin V2robbyczgw-cla/web-search-plus-plugin101—~1.9kAutomated safety check: NotesMIT

Similar skills

  • Ultimate Search

    ckckck/UltimateSearchSkill

    双引擎网络搜索:Grok AI 搜索(实时联网+AI分析)+ Tavily 搜索(结构化结果+网页抓取). An agent skill from ckckck/UltimateSearchSkill.

    289 GitHub stars~944 tokensUpdated 6 mo ago
    Productivity & AutomationAuto-check: notes
  • Web Search

    EXboys/skilllite

    Web search and content extraction with Tavily and Exa via inference.sh CLI.

    170 GitHub starsUsed in 2 repos~1k tokens
    Productivity & AutomationAuto-check passed
  • Mysearch

    skernelx/MySearch-Proxy

    Install, verify, debug, and use MySearch MCP/Skill. An agent skill from skernelx/MySearch-Proxy.

    159 GitHub stars~3k tokensUpdated 6 mo ago
    Productivity & AutomationAuto-check: notes
  • Tavily

    openclaw/openclaw

    Tavily web search, content extraction, and research tools. An agent skill from openclaw/openclaw.

    392k GitHub starsUsed in 2 repos~1.2k tokens
    Productivity & AutomationAuto-check passed
  • Web Search Plus Plugin V2

    robbyczgw-cla/web-search-plus-plugin

    OpenClaw plugin for source-only Routing v2 multi-provider search, completion-order Research with quality quorum and attributed provenance, heading-aware extraction spans, Tavily-first extraction…

    101 GitHub stars~1.9k tokensUpdated 16 days ago
    Productivity & AutomationAuto-check: notes
  • Tavily Hikari Best Practices

    IvanLi-CN/tavily-hikari

    Guide agents to use Tavily Hikari safely through tvly-hikari.

    362 GitHub stars~287 tokensUpdated 2 days ago
    Productivity & AutomationAuto-check passed

More from Mathews-Tom/armory

All 80 skills in this repo
  • Architecture Reviewer

    Mathews-Tom/armory

    Architecture reviews across 7 dimensions (structural, scalability, enterprise readiness, performance, security, ops, data) with scored reports.

    328 GitHub stars~4.6k tokensUpdated 4 days ago
    Auto-check passed
  • Concept To Image

    Mathews-Tom/armory

    Turn concepts into static HTML visuals exported as PNG or SVG files via HTML/CSS/SVG.

    328 GitHub stars~2.6k tokensUpdated 4 days ago
    Auto-check passed
  • Watch

    Mathews-Tom/armory

    A skill your agent uses when analyzing an existing video URL or local recording: "watch this video", "analyze youtube video", "summarize this video", "youtube transcript", "find this moment", "what…

    328 GitHub stars~2.8k tokensUpdated 4 days ago
    Auto-check passed
  • Code Refiner

    Mathews-Tom/armory

    Deep code simplification and refactoring preserving behavior across Python, Go, TypeScript, Rust.

    328 GitHub stars~3.1k tokensUpdated 4 days ago
    Auto-check passed
  • Concept To Video

    Mathews-Tom/armory

    Turn concepts into animated explainer videos using Manim (Python) with MP4/GIF output, audio overlay, multi-scene composition.

    328 GitHub stars~4.9k tokensUpdated 4 days ago
    Auto-check passed
  • Decision Map

    Mathews-Tom/armory

    Maps the unresolved architecture, policy, and scope decisions that must be answered before planning can start: one durable decision ticket per question on the issue tracker, typed and blocker-linked…

    328 GitHub stars~2.7k tokensUpdated 4 days ago
    Auto-check passed

Works with

Questions about Tavily

What does Tavily do?

AI-optimized web search via Tavily API when no URL is provided. Tavily is an agent skill from Mathews-Tom/armory. AI-optimized web search via Tavily API when no URL is provided.

When should I use Tavily?

Tavily fits situations like: : search the web; find recent information.

How do I install Tavily in Claude Code?

Run `npx skills add Mathews-Tom/armory --skill tavily -a claude-code`. Or copy the skill folder (skills/tavily in Mathews-Tom/armory) into .claude/skills/tavily in your project. Claude Code loads it when a task matches its description.

How do I install Tavily in Codex?

Run `npx skills add Mathews-Tom/armory --skill tavily -a codex`. Or copy the skill folder (skills/tavily in Mathews-Tom/armory) into .agents/skills/tavily in your project. Codex loads it when a task matches its description.

Can I use Tavily in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Mathews-Tom/armory --skill tavily -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/tavily, .gemini/skills/tavily, .github/skills/tavily and .opencode/skills/tavily in your project.

What does Tavily need to run?

Going by SKILL.md and its folder, Tavily needs JavaScript for the scripts in its folder, the command-line tools its instructions call (node) and credentials named TAVILY_API_KEY. Our summary lists: Node.js; A credential in TAVILY_API_KEY.

Does Tavily access the network?

SKILL.md names 2 domains. In commands or code: source.com; the agent is likely to contact it when it follows the instructions. As links in the text: tavily.com. This is read from the text; nothing was executed.

Is Tavily safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Tavily use?

Tavily is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Tavily use?

About 1.5k tokens (SKILL.md is roughly 6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Tavily?

Skills that share tags, products or a category with Tavily: Ultimate Search (ckckck/UltimateSearchSkill, 289 stars), Web Search (EXboys/skilllite, 170 stars), Mysearch (skernelx/MySearch-Proxy, 159 stars) and Tavily (openclaw/openclaw, 392k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Tavily?

Mathews-Tom (a GitHub user) maintains it in Mathews-Tom/armory, which has 328 GitHub stars. The repository holds 80 skills in this directory. The repository was last updated on October 6, 2026.

Source: Mathews-Tom/armory on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.