Scrape
Extract markdown from one or more URLs.
marmot scrape <url> [<url> …] [flags…]Providers
exa, firecrawl, parallel, tavily.
Flags
| Flag | Description |
|---|---|
--provider <slug> | One of the four. Falls back to defaults.scrape.provider. |
--api-key <key> | Override the env var for this call. |
--format <markdown|text|html> | Output format. Provider may return a subset. Default markdown. |
--query <text> | Optional intent for chunk reranking (Tavily). |
--raw | Emit the provider's native response under raw. |
--json | Emit the structured envelope (default). |
--no-cache | Bypass the response cache for this call. Skip read and skip write. |
--refresh | Skip cache read but write the fresh response, overwriting any cached entry. |
--retries <n> | Retry retryable provider errors up to N times with exponential backoff. Default 0, max 10. |
--timeout <seconds> | Per-attempt request timeout. Default 120. |
Cache flags only apply when caching is enabled for this provider via providers.<slug>.cache.enabled. See Caching.
Presets
scrape-mode presets accept urls (list — appends with runtime), format, query, cache, refresh, output, raw, retries, timeout, session. New negation flags: --cache, --no-refresh, --no-raw. See Presets — Merge rules.
marmot preset create md-scrape --mode scrape --provider firecrawl --format markdown
marmot @md-scrape https://example.com https://other.comConfig keys
{ "defaults": { "scrape": { "provider": "firecrawl" } } }