Tavily Search Pro

v1.0.0

Tavily AI search platform with 5 modes: Search (web/news/finance), Extract (URL content), Crawl (website crawling), Map (sitemap discovery), and Research (deep research with citations). Use for: web search with LLM answers, content extraction, site crawling, deep research.

6· 7.7k·38 current·38 all-time
byshaharsh@shaharsha
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (web search, extract, crawl, research) align with requested resources: only a single TAVILY_API_KEY is required and the code posts requests to https://api.tavily.com. Nothing in the manifest requests unrelated cloud credentials or system access.
Instruction Scope
Runtime instructions direct the agent/user to run the included Python script which sends queries and URL lists to the Tavily API for search, extraction, crawling and research. This is expected for a remote search service, but it means any URLs, page content, or queries you pass will be transmitted to a third-party API (Tavily).
Install Mechanism
Install is via pip (tavily-python) invoked by the included install.sh; this is a typical mechanism but carries standard package-install risks (unreviewed PyPI package). Minor inconsistency: registry metadata reported 'No install spec' but SKILL.md contains an install entry and an install.sh is present. No downloads from unknown hosts or archive extraction were observed.
Credentials
Only TAVILY_API_KEY is required (declared as primary). That key is used in Authorization headers to the Tavily API — appropriate for this service. No unrelated secrets or multiple credentials are requested.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or global agent settings. It is user-invocable and can be invoked autonomously (platform default) — no extra persistent privileges were requested.
Assessment
This skill behaves as a client for the Tavily API and will send your search queries and any URLs you ask it to extract or crawl to https://api.tavily.com using the TAVILY_API_KEY you provide. Before installing: (1) Confirm you trust Tavily to receive the content you will send (do not submit sensitive URLs, credentials, or unpublished documents). (2) Inspect the tavily-python package on PyPI (or the package source) to ensure it is legitimate. (3) Note the minor metadata mismatch: SKILL.md includes an install step and there is an install.sh — the registry entry omitted an install spec. (4) If you must audit network behaviour, run the script in a controlled environment and monitor outbound requests. If you are unsure about trusting a third party with crawled/extracted content, do not install or restrict usage to non-sensitive queries only.

Like a lobster shell, security has layers — review code before you run it.

latestvk979rtf422kech181jzmecv91n80qvx3

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

🔎 Clawdis
EnvTAVILY_API_KEY
Primary envTAVILY_API_KEY

Comments