LLMSkillHub
mirroredCopied from 7df-lab/devo under MIT. 7df-lab has not claimed this page, and is not affiliated with LLM SkillHub.

7df-lab / websearch

Search the web. Prefer Devo hosted web_search (provider-hosted tool or local Exa/Tavily). Fall back to Serper only when hosted search is unavailable. Use when the user asks to search the web and no web_search tool is already available.

0
SKILL.md
1.7 KB

Web Search

Prefer Devo's built-in search. Use this skill only when you cannot call a
model-facing web_search tool directly.

Preference order

  1. Provider-hosted or local web_search tool — if your tool list includes
    web_search, call that tool. Do not route through this skill.

  2. This skill — tries host web.search (local Exa / Tavily when configured),
    then Serper if a key is saved.

  3. If nothing is configured — tell the user how to enable search (below).
    Do not invent results.

Enable hosted search (preferred)

  • Provider-hosted: set [tools.web_search] mode = "provider" and use a
    model on Anthropic Messages or OpenAI Responses that supports hosted search.

  • Local Exa / Tavily: set [tools.web_search] mode = "local" and configure
    a provider under [tools.web_search.local_providers.<id>] with
    kind = "exa" or kind = "tavily" (credentials via /login / auth.json).

Serper fallback (third party)

Only when hosted search is not available:

  1. Get a free API key at https://serper.dev

  2. In Devo, run /login → MCP Connections → Serper (web search) and
    paste the key.

Do not ask the user to set environment variables.

Optional overrides:

  • DEVO_WEBSEARCH_TIMEOUT — HTTP timeout in seconds (default 45).
  • DEVO_WEBSEARCH_NUM_RESULTS — organic results to return (default 5).

Usage

print(await websearch.run("latest Devo release"))