Adapter: Pixabay #1409

👋 A self-contained first contribution: one JSON adapter file, a spec and a logo. No need to know the rest of the codebase.

Why this one

In the last 30 days, 2 different workspaces on AnythingMCP Cloud searched the catalog for Pixabay and found nothing. People want it; nobody has built it yet.

The API

• Docs: https://pixabay.com/api/docs/
• Auth: a free API key in the key query parameter
• Category: Media

Suggested first tools (read-only)

search images, search videos (with filters for type, orientation, category, safesearch). Read-only tools are enough for a first version; write tools can follow in a later PR.

How

Follow the six steps in AGENTS.md → Add an adapter and docs/tool-definition.md. The scaffold command for this one:

npm run adapter:new -- pixabay --region intl --auth QUERY_AUTH

Then node scripts/validate-adapters.mjs --warn, node scripts/regenerate-catalog.mjs, a intl/pixabay.live.spec.ts (static assertions always, live calls behind RUN_PIXABAY_LIVE=1), node scripts/adapter-count.mjs --check and the logo. The walkthrough and past examples are in #982.

Done when

• The adapter validates without new warnings and the catalog is regenerated.
• Every tool has a description an AI can act on (what it returns, when to use it) and a response mapping that keeps the useful fields.
• The PR says which calls you ran against a real account, or that you could not and built it from the docs.

Claim it

Comment here to claim it. A claim without a PR for 14 days goes back to the pool, no hard feelings. Sign the CLA when the bot asks.