A walkthrough of how Slackscraper actually works, from the OAuth handshake to the parquet file landing in your bucket. No magic, just plumbing.
OAuth handshake with your workspace. We request the minimum read scopes Slack supports: channels:history, groups:history, im:history, mpim:history, files:read, reactions:read. A workspace admin approves and we never ask for write access — we cannot send messages, modify channels, or impersonate users.
Slackscraper crawls every channel you authorize, in parallel, respecting rate limits. Backfill is resumable: if the connection drops, it picks up where it left off. Idempotent: re-running it never duplicates a message.
After backfill, Slackscraper subscribes to your workspace's event stream over Slack's Events API. New messages, edits, deletions, reactions, and file uploads land in your warehouse within 200ms of being sent. Hourly and daily modes available if you don't want streaming.
You configure where the data lands — your S3 bucket, your Snowflake account, your Postgres instance, or a flat CSV on your laptop. We hold sync state, but the data itself goes directly to you. Nothing of yours persists on our infra (on Team) past the moment it transits.
We cannot send messages, modify channels, or impersonate users. Period.
TLS 1.3 from Slack to us. TLS 1.3 from us to your destination. Always.
On Team, messages stream directly to your bucket. We hold sync state only.
An admin can remove the app at any time. We stop receiving events immediately.
Audit in progress. Q3 2026 completion. SOC 2 Type I available on request.
EU-region buckets and a signed Data Processing Agreement on request.
BAA available for Team customers. We do not process PHI on Free or Individual.
Every sync, every config change, every API call — logged and exportable.
Free tier, no card. Install on Slack, run one sync, see what comes out.
Add to Slack →Built in Brooklyn + Lisbon.
SOC 2 in progress. EU residency available.