Appearance
anti-coordination
FreshSource: BIBLE.md (Phase 12 Phase 4, anti-coordination 4 routes), ROADMAP.md (Phase 4.2)
Network graph that tracks which accounts exist on which platforms and ensures they NEVER interact, never share IPs, never overlap in timing.
Not just randomization. Deliberate anti-correlation. If Account A and Account B are both on Twitter, they never follow each other, never engage with the same content on the same day, never use the same proxy.
What it detects
flowchart LR A[Account A] --> P1[Proxy 1.2.3.4] B[Account B] --> P1 P1 --> CN[CONFLICT: shared proxy] C[Account C] --> T[Twitter @somepost at 12:01] D[Account D] --> T T --> CT[CONFLICT: shared content engagement]
Conflict types:
- Shared proxy across accounts
- Same content engaged within a window (likes, comments, follows)
- Timing overlap on the same platform within a window
- Mutual follow/connect
Routes (4)
| Method | Route | What |
|---|---|---|
| GET | /anti-coordination/graph | Full network graph |
| GET | /anti-coordination/conflicts | Detected conflicts |
| POST | /anti-coordination/check | Pre-flight check before action |
| POST | /anti-coordination/resolve | Auto-resolve a conflict |
Pre-flight check
Before any action:
bash
curl -X POST http://localhost:4700/anti-coordination/check \
-H "Content-Type: application/json" \
-d '{
"accountId": "abc123",
"platform": "twitter",
"action": "like",
"target": "tweet-id-456"
}'Returns:
json
{
"safe": false,
"conflict": {
"type": "shared-content",
"withAccount": "def789",
"engagedAt": "2026-05-20T11:59:00Z",
"windowMinutes": 60
},
"recommendation": "delay-or-skip"
}Related
- behavioral-identity: per-account fingerprints
- geofenced-personas: geo-coherence
- canary-system: platform threshold detection