Public bid books.
Public pages and API responses where a network lists its current bids, leaderboard, or advertisers.
Research crawler
The PicassoAdIntel crawler.
Run by Picasso. What it reads, how it behaves, and how to stop it.
picasso / ad-intelindexed
Research crawler
The PicassoAdIntel crawler
Run by Picasso. What it reads, how it behaves, and how to stop it.
Public bid booksPublic pages and API responses where a network lists its current bids, leaderboard, or advertisers.
Advertiser catalogsPublic advertiser lists, storefront pages, and sitemaps.
Transparency filesThe standard sellers.json, ads.txt, and app-ads.txt files, which sites publish so that others can read them.
PicassoAdIntel/1 (+https://picassohq.com/ad-intel)User-agent: PicassoAdIntelDisallow: /What it is.
PicassoAdIntel/1 is a research crawler operated by Picasso. It reads advertising listings that sites post publicly, so Picasso can see which companies advertise to developers inside AI tools. This page covers only requests that carry the User-Agent below.
PicassoAdIntel/1 (+https://picassohq.com/ad-intel)What it reads.
Pages and files that sites post publicly for anyone to read.
Advertiser catalogs.
Public advertiser lists, storefront pages, and sitemaps.
Transparency files.
The standard sellers.json, ads.txt, and app-ads.txt files, which sites publish so that others can read them.
What it requests.
robots.txt first.
It reads a host's robots.txt before anything else on that host. It follows the group for the PicassoAdIntel token, or the group for * when no group names it, and it does not request a path those rules disallow.
When robots.txt is unreadable.
If robots.txt answers 401 or 403, or the request for it fails, it treats the whole site as disallowed and requests nothing else from it.
The only exception.
A path is read despite a robots.txt rule only when Picasso's owner has approved it in a dated, expiring written record that names what it covers.
One request a second, at most.
It waits at least one second between requests to the same host and sends them one at a time.
A plain User-Agent.
Every request identifies itself as PicassoAdIntel/1 and links to this page.
What it never does.
These limits apply to every request that carries the PicassoAdIntel/1 User-Agent.
No sign-ins.
It does not log in to a site, hold a session there, or send cookies. The only credential it presents anywhere is a read-only token for GitHub's public code search API, which GitHub requires.
No purchases or payouts.
It does not buy anything, accept an offer, or claim a reward. A deny list refuses checkout, purchase, payout, earnings, claim, and accept paths before any request is sent.
No clicks, impressions, or billing.
It does not request ad-serving, click, impression, tracking, or beacon endpoints, so reading a listing cannot count as a view, a click, or a billable event. The deny list is checked again on every redirect.
No personal data stored.
Email addresses, phone numbers, user names, IP addresses, and wallet addresses that appear in a listing are discarded before anything is stored.
How to opt out.
Add a rule for the PicassoAdIntel token to your robots.txt. The crawler picks up the change within a day and stops requesting the paths you list. To block your whole site:
User-agent: PicassoAdIntelDisallow: /Block one section.
List its path instead of /, for example Disallow: /listings/.
Your existing rules apply.
If no group names PicassoAdIntel, the crawler follows the group for *.
Ask by email.
Write to gchahal@auraone.ai and name the host. Picasso will stop requesting it.
Contact.
This page is maintained by Picasso. Questions, corrections, and opt-out requests go to gchahal@auraone.ai.