Anti-spam — suppression, stop-list, marking, block-guard

TL;DR for agents: Before each send the pipeline checks: 1) is the recipient email in suppression? 2) is the recipient domain/company in stop-lists? 3) does the rendered body trigger any marking/patterns regex? Any hit blocks the send and logs the reason. After sending, 4) block-guard watches provider blocks and auto-pauses accounts / barks templates — inspect it at mailing/block-guard/dashboard.
Plain-text mirrors: /llms.txt · /anti-spam.md

Suppress an email globally

Scope: suppression:write. Use this when a recipient asks to never be contacted again — bounces and unsubscribes also auto-fill this list.

curl -s -X POST "https://api.live-direct-marketing.online/api/suppression" \
  -H "Authorization: Bearer $LDM_KEY" -H "Content-Type: application/json" \
  -d '{"email":"do-not-contact@example.com","reason":"unsubscribed"}'

Check before send

Scope: suppression:read. The pipeline calls this for every recipient — agents can call it manually for pre-flight checks.

curl -s -X POST "https://api.live-direct-marketing.online/api/suppression/check" \
  -H "Authorization: Bearer $LDM_KEY" -H "Content-Type: application/json" \
  -d '{"emails":["a@example.com","b@example.com"]}'

Stop-list a whole company / domain

Scope: stoplist:write. For competitor/sensitive domains. Stops outreach to any recipient under that domain.

curl -s -X POST "https://api.live-direct-marketing.online/api/stop-lists" \
  -H "Authorization: Bearer $LDM_KEY" -H "Content-Type: application/json" \
  -d '{"value":"competitor.com","type":"DOMAIN","reason":"competitor"}'

Marking pattern (auto-classify outbound content)

Scope: marking:write. Regex matched against subject+body. When triggered, the dialog is auto-tagged for review. Used for compliance keywords.

curl -s -X POST "https://api.live-direct-marketing.online/api/marking/patterns" \
  -H "Authorization: Bearer $LDM_KEY" -H "Content-Type: application/json" \
  -d '{"pattern":"(?i)(crypto|forex|gambling)","label":"sensitive","action":"block"}'

BlockGuard — reactive provider-block defense

Scope: mailing:read (SUPER/MANAGER). The first three layers run before a send; BlockGuard runs after. It records every provider block, then auto-pauses an account (3 blocks/day) or quarantines ("barks") a template (10 blocks from 3+ accounts across 2+ providers within 7 days). A 1-hour cooldown stops one bad batch from barking a whole template bank at once.

curl -s "https://api.live-direct-marketing.online/api/mailing/block-guard/dashboard" \
  -H "Authorization: Bearer $LDM_KEY"

Drill into who is burning: block-guard/providers (per-provider totals), block-guard/account/:accountId and block-guard/template/:templateId (block-event history). To release a template BlockGuard barked, scope mailing:write:

curl -s -X POST "https://api.live-direct-marketing.online/api/mailing/block-guard/template/$TEMPLATE_ID/unblock" \
  -H "Authorization: Bearer $LDM_KEY" -H "Content-Type: application/json" \
  -d '{"creativeId":"<creative-id>"}'