Vai al contenuto
TakoAPI
Torna alle skill

kagi-enrich

di joelazar

Web & Frontend Development
344Download riportati da clawskills.sh
0Visualizzazioni
0 likeAccedi per mettere like a questa skill
—Non ancora valutato
La tua valutazione

Sintesi

Search Kagi's unique non-commercial web (Teclis) and non-mainstream news (TinyGem) indexes for independent, ad-free.

Descrizione

Cosa fa questa skill

Searches Kagi's proprietary Teclis (independent web) and TinyGem (non-mainstream news) indexes via the Kagi Enrichment API. Returns results from personal blogs, indie projects, niche outlets, and sources that mainstream search engines bury under commercial content. Uses a Go binary with no runtime dependencies.

Unlike general search, both indexes are curated specifically to exclude commercial and SEO-optimized content, so results surface independent voices that mainstream engines deprioritize.

Quando usarla

  • Finding independent blog posts on a technical topic
  • Discovering niche journalism outside major news outlets
  • Researching open-source projects without SEO noise
  • Getting non-commercial perspectives on a product or technology
  • Surfacing Hacker News threads and Reddit discussions on a topic

Esempio di workflow

INPUT

User asks: find independent writing about SQLite internals

AGENT
  1. 1

    Run kagi-enrich.sh with the web index and query 'sqlite internals'

  2. 2

    Receive ranked results from Teclis index with titles, URLs, snippets, and publish dates

  3. 3

    Filter or limit results with -n flag if needed

  4. 4

    Optionally re-run with news index to find related discussions

  5. 5

    Return results with source URLs and snippets to user

OUTPUT

List of independent blog posts and technical write-ups about SQLite internals with URLs, snippets, and API balance remaining

Requisiti

  • KAGI_API_KEY environment variable
  • Kagi account with API access enabled (kagi.com/signup)
  • Kagi API credit balance (billing at kagi.com/settings/billing_api)

Installazione

clawhub install joelazar/kagi-enrich

Esegui il comando CLI oppure incolla il prompt in Claude Code, Cursor, Codex o nell'agente di coding che usi.

Link