grep.boolsai.ai

Parallel regex search across every scan we've ever taken. Fanned out across R2, ~10K-200K objects per query.

Endpoints

GET /grep?pattern=… — streams NDJSON of matches

GET /sample?pattern=…&limit=20 — JSON array of first N matches

GET /count?pattern=… — count only, no snippets

Filters

• pattern — JS regex (case-insensitive by default)
• limit — max matches (default 200, max 2000)
• max_scan — max R2 objects to read (default 10000, max 200000)
• tld — e.g. com.au · since=2026-05-01 · domain_like=patagonia
• vendor — pre-narrow via id_index (e.g. vendor=shopify)
• case_sensitive=1 — opt-in

Examples

curl "https://grep.boolsai.ai/sample?pattern=window\._fs_org&limit=10"
curl "https://grep.boolsai.ai/count?pattern=cdn\.optimizely\.com&vendor=shopify"
curl "https://grep.boolsai.ai/grep?pattern=mparticle.*data_plan&limit=50"