generated: '2026-10-07'
method: searched
source: https://s1.dev/docs/integrations/cli; https://raw.githubusercontent.com/superagents-lab/search1api-cli/master/README.md
description: 's1, the first-party Search1API CLI: OAuth 2.1 login (dynamic client registration + PKCE), web/news
  search, page crawl, sitemap, trending, balance, and s1 learn which turns a URL into an installable Agent Skill.
  Prints JSON with --json for scripting.'
docs: https://s1.dev/docs/integrations/cli
repo: https://github.com/superagents-lab/search1api-cli
install:
  script: curl -fsSL https://cli.search1api.com/install.sh | bash
  npm: npm install -g search1api-cli
commands:
  auth_config:
  - name: login
    description: Authorize with OAuth 2.1 in the browser (dynamic registration, Authorization Code + PKCE S256);
      --no-browser prints the URL.
  - name: config set-key | show
    description: Manage configuration; set-key stores an API key fallback, show reports the active credential source.
  - name: update
    description: Update s1 in place.
  search:
  - name: search "<query>"
    description: Search the web (POST /search).
    flags:
    - -n/--max-results 1-50 (default 10)
    - -s/--service engine (default google)
    - -c/--crawl N crawl the top N results for full text
    - --include/--exclude sites
    - -t/--time day|month|year
    - --json
  - name: news "<query>"
    description: Search the news (POST /news).
  - name: trending <github|hackernews>
    description: Trending topics (POST /trending).
  retrieval:
  - name: crawl <url>
    description: Read a page (POST /crawl).
  - name: sitemap <url>
    description: List a site's links (POST /sitemap).
  account:
  - name: balance
    description: Remaining credits (GET /usage).
  skills:
  - name: learn
    description: Turn a URL (or a documentation site with --site) into an installable Agent Skill; added in 1.4.0.
key_flows:
  deep_search:
    command: s1 search "web framework" -c 3
    description: Search, then crawl the top 3 results for their full text; bills 1 credit for the search plus 1
      per page crawled.
  scripting:
    command: s1 search "test" --json | jq '.results[0].title'
    description: Machine-readable output straight into jq.
env:
- SEARCH1API_KEY takes precedence over saved OAuth credentials
removed: s1 reasoning and s1 models were removed in CLI 1.2.3 (endpoint discontinued).