How to let your agent find tools via the Botlens API

Start with the manifest

GET https://botlens.pages.dev/api/v1/index.json

It lists badge counts, how many search shards exist, and the row format: [id, name, badge, category, coverage, source_index, author, description].

Search the shards

Safe and Caution rows live in /api/v1/search/SC-0.json and onward; unscanned ones in U-*.json. Each shard holds up to 2,000 rows. Filter by name and description, prefer badge S, then C.

Fetch the full record

GET https://botlens.pages.dev/api/v1/bots/{floor(id/250)}.json

The record has the scan findings, coverage, commit, SHA-256 and source URL. Show the findings to your user before installing.

Machine-readable docs

openapi.json and llms.txt describe the same thing for agents. A hosted search endpoint and a submission API are planned; see For agents.

Last checked: 2026-11-01 · How Botlens checks listings: a rule-based scan of public instructions and code, with coverage recorded per listing.

The Botlens API is a free API to find AI tools for agents. It is plain static JSON with no key. An agent reads one manifest, searches shards of up to 2,000 rows each, prefers Safe over Caution, and fetches a full record with scan findings to show its user before installing.

Is there an API to find AI tools for agents?

Yes. Botlens publishes its catalog as static JSON under /api/v1/. There is no key and no rate plan to sign up for. The manifest at /api/v1/index.json gives badge counts, how many shards exist and the row format. Today it lists 30,757 listings, with 7 Safe-and-Caution shards and 10 Unscanned shards.

Each search row has eight fields: id, name, badge, category, coverage, source index, author and a short description. That is enough for an agent to filter by task and safety without loading a web page.

How should an agent use the results?

An agent should rank by fit first, then by badge and coverage. A good order is Safe with full code, then Safe with less coverage, then Caution. Unscanned rows live in separate U-*.json shards, so an agent can leave them out by default.

Before installing, the agent should fetch the full record from /api/v1/bots/{floor(id/250)}.json. The record includes the scan findings, coverage, commit, SHA-256 and source URL. Show those findings to the user and ask before installing. Risky listings are not in the API at all.

What is llms.txt for agents?

llms.txt is a plain text file at the site root that tells AI agents and AI search tools what a site is and where the useful parts are. The Botlens llms.txt describes the catalog, the general checking method, the key guides and every category page with its counts. An agent that reads it first knows to use the JSON API instead of scraping HTML.

Botlens also ships openapi.json, which describes the same endpoints in a format many agent frameworks can load directly.

Do I need to build this myself?

No, not if you use Grok Bot. The Botlens skill already does this search for you.

Let Botlens pick for you

Yes. Botlens is a free catalog of Grok Bot bots, templates and skills that shows each listing's safety status, and it ships a Grok Bot skill that picks them for you. Describe your task and the skill searches the catalog, skips listings flagged Risky, ranks matches by fit and safety, and asks before installing anything. You don't have to test bots by hand.

See How to find the right AI bots and skills automatically or install the skill.

FAQ

Do I need an API key?

No. The read API is static JSON.

Are Risky bots in the API?

No. They are excluded everywhere.