Slack to Arcads Automation: 8 Production Checks
FlowPatchA Slack-to-Arcads automation looks simple until retries, stale requests, missing approvals, or depleted credits turn one brief into duplicate jobs. This checklist defines a narrow production boundary for teams that want Slack to collect an approved UGC request and Arcads to generate exactly one tracked result.
Independent note: FlowPatch is not affiliated with Arcads or Slack. This is an implementation checklist, not a customer case study.
1. Confirm API and credit readiness
Arcads' official API guide says credentials are generated under Settings -> Public API and uses a client ID plus client secret. Its official MCP guide also warns that generation requires an active subscription with remaining credits. Verify access before building any Slack interaction.
- The client owns the Arcads workspace, credentials, plan, and generation credits.
- Credentials stay in server-side secret storage and never enter a Slack modal or channel.
- One approved actor, template, or generation recipe is named in scope.
2. Make Slack collect a production contract
Use a modal built from Block Kit inputs for structured fields: client or brand, approved script, output format, actor or template, requester, approval state, and destination channel or thread. Slack's official modal documentation recommends unique block_id and action_id values so submitted fields can be identified reliably.
3. Verify Slack before parsing
Slack signs inbound requests. Compute the documented HMAC from the raw request body and timestamp, use a timing-safe comparison, and reject requests older than five minutes. A failure here must create no Arcads job.
4. Acknowledge within three seconds
Slack requires valid interaction payloads to be acknowledged within three seconds. Return the acknowledgment first; run the Arcads submission and polling work asynchronously. This prevents a long generation request from appearing as a failed Slack command.
5. Validate before consuming credits
- Require every agreed field and explicit approval state.
- Check script length, URL shape, and allowed output options.
- Confirm the selected generation recipe is on the allowlist.
- Reject missing rights or claims approval before submission.
6. Prevent duplicate generations
Derive an idempotency key from the Slack interaction or request ID. Persist it before calling Arcads. When Slack retries the same request, return the existing internal job and Arcads run ID rather than creating another credit-consuming generation.
7. Track one asynchronous Arcads run
Arcads documents a script flow that creates products, folders, and scripts, starts generation, then retrieves video status and the download link. Store the external run identifier, timestamps, bounded retry count, and terminal state. Time out deliberately instead of polling forever.
8. Return a result or actionable failure
Post the finished asset link to the originating Slack context. For quota, validation, timeout, or vendor errors, return a human-readable state plus a safe correlation ID - never a token, client secret, raw authorization header, or unnecessary script content.
Minimum acceptance tests
- Valid signed input opens or submits the agreed flow.
- Bad signatures and stale timestamps are rejected.
- Missing fields create no Arcads generation.
- The same Slack interaction cannot create two jobs.
- A successful run posts one usable result link.
- Failure and timeout states expose no credentials.
- The run log contains request ID, Arcads run ID, status, timestamps, and result URL.
Try the boundary before building
Use the safe Slack-Arcads simulation to inspect request, approval, duplicate-protection, generation, and result states without connecting credentials or consuming credits. Then generate a private Arcads API pilot scope or review the fixed $299 Slack-Arcads pilot.