Docs
Getting started with SourceFinch
SourceFinch reads a public page, feed or API on a schedule and shows it as rows. Each time, it tells you which rows and fields changed, and keeps a dated copy of what it read. This page takes you from a new account to your first change alert. To see it work first, try a page on the homepage; no account is needed.
1. Create your account and workspace
Sign up with your email (or Google, where enabled) and confirm the link we send. Then name your workspace and accept the Terms and Acceptable Use Policy. New workspaces start with 14 days of Pro, with no card required, then continue on Free unless you upgrade.
2. Add a source
Choose New source, paste a public https:// address and press Inspect. SourceFinch reads the page once and looks for things that repeat: table rows, list items, feed entries, items in a data file. Pick the group you want, rename or drop fields, and check the preview. Or start from a Verified source or a template.
- Only public HTTPS addresses are read. Private and internal networks are refused.
- robots.txt is followed on every page, every time.
- Pages that build their content with JavaScript need browser rendering, which is on Pro and above. Free reads pages, feeds and APIs that load without a browser.
- A new website is approved automatically on first use unless it is on our blocklist; Free workspaces can add up to 3 new websites a day.
Choose a key field: something that names each row, such as an ID or a link. With a key, an edited row shows up as an edit, field by field, instead of one row removed and another added.
3. Run it once, then put it on a schedule
Save the source and run it. The first run that passes its checks is your baseline: SourceFinch keeps a dated copy of what the site served and the records read from it. Then set a schedule: weekly, daily, hourly on paid plans, or every 15 minutes for HTTP sources on Business. The picker shows how many runs the schedule uses of your monthly allowance.
4. See what changed, and the proof
Every later run is compared with the last one that passed its checks. Changes lists the rows added, removed and edited across your sources; open a run to see the before and after of each edited field. Select any row to open the copy it was read from, with the time it was read and the address it came from. How evidence works explains what is kept and how to check it yourself.
5. Alerts and team
In Alerts, choose an email for each change, a daily or weekly digest, or none, and whether to hear about failures and usage. Each source can have its own setting. A change email lists what changed (before → after) and links to the copy each value was read from. Every email has an unsubscribe link. In Team, owners invite editors and viewers up to the plan’s seat limit.
6. Export and API
Export any source’s current records as CSV or JSON. For programmatic access, create a project API key in API & access (read or read/write scope; the secret is shown once) and call the REST API:
curl https://sourcefinch.com/v1/sources/<source id>/records \ -H "Authorization: Bearer $SOURCEFINCH_API_KEY"
Every value comes with the ids of the copies it was read from. The API reference lists every operation, and the OpenAPI 3.1 description is at /v1/openapi.json. Errors are problem details with a stable code (plan limits are HTTP 402 with code: "plan_limit"). The older action endpoint, POST /api/v1, still works but takes no new actions.
Developer tools on the same API: the TypeScript SDK, the sourcefinch CLI and the MCP server for AI assistants.
7. Plans, limits and billing
Plan & usage shows runs, sources, evidence storage and API keys against your plan. Owners upgrade through Stripe checkout and manage invoices, payment methods and cancellation in the Stripe billing portal. See pricing for plan limits.
Questions? Email hello@sourcefinch.com.