Tavily
TavilySearch API shaped for agents: results plus a short synthesised answer.
Score a search result set for relevance before an agent trusts it.
Agents that call search and need to know whether the results are any good
Judging the search engine itself — it scores one result set, at one moment
Given a query and the returned results, it labels each result relevant, partial, or off-topic, with a one-line reason. The pattern that matters is the refusal: if fewer than two results are relevant, the skill reports a weak result set instead of letting the agent synthesise from noise.
Clone the repository, then copy the skill folder into your agent's skills directory — ~/.claude/skills/ for Claude Code. Skills are plain files: no build step, no daemon.
npx skills add search-result-evalRead the SKILL.md before you install. A good skill states when it applies, what it must not do, and how to verify the result. A skill that only lists adjectives will change the agent's tone without improving its output — skip those.
Search API shaped for agents: results plus a short synthesised answer.
Enterprise search and an assistant that answers from your company's own documents.
Crawl and convert pages to clean markdown for retrieval.
A community server for the Pexels API: search photos and videos by query, orientation, and size. It solves the image problem for content templates where a generic ph…
A community server for Unsplash search and photo metadata. Unsplash's strength is editorial photography rather than product shots, which makes it the better fit for …
Adds web and local search through the Brave Search API, returning titles, snippets, and URLs the agent can then fetch. Requires your own API key. Archived, but still…