mirror of
https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git
synced 2026-06-18 07:19:35 +00:00
20277e0afc
8 agents + 4 models + 4 configs + CLI dry-run + 3 docs. Tested on agency/real_estate/clinic/saas — all pass. Safety: LinkedIn scraping PROHIBITED, WhatsApp blast PROHIBITED. https://claude.ai/code/session_01W1rJthWDkasijTdXCfxVHs
67 lines
2.5 KiB
Markdown
67 lines
2.5 KiB
Markdown
# Dealix GTM OS — Architecture
|
|
|
|
## Overview
|
|
```
|
|
Company Input → Research → Enrichment → ICP Detection → Opportunity Mapping
|
|
→ Channel Strategy → Message Generation → Compliance Check
|
|
→ Human Approval / Safe Automation → CRM Tracking → Learning Loop
|
|
```
|
|
|
|
## Layers
|
|
|
|
### A. Data Layer
|
|
Collects from allowed sources only:
|
|
- Uploaded company files
|
|
- Public websites
|
|
- Google Programmable Search / Tavily
|
|
- Official APIs
|
|
- CRM data
|
|
- Inbound messages
|
|
- Manual imports
|
|
|
|
### B. Intelligence Layer
|
|
13 specialized agents understand companies, markets, and opportunities.
|
|
|
|
### C. Compliance Layer
|
|
Decides: allowed / manual_required / opt_in_required / prohibited
|
|
|
|
### D. Execution Layer
|
|
Only safe actions: drafts, CRM tasks, scorecards, content packs, approved campaigns.
|
|
|
|
### E. Learning Layer
|
|
Tracks replies, demos, conversions. Updates ICP, scoring, messages, channels weekly.
|
|
|
|
## 13 Agents
|
|
|
|
| Agent | Role | Input | Output |
|
|
|-------|------|-------|--------|
|
|
| Supervisor | Orchestrates all | CompanyInput | Full GTM Pack |
|
|
| Company Research | Understands company | Name/URL/sector | CompanyIntelligence |
|
|
| Web Search | Searches allowed sources | Query | SearchResults |
|
|
| Enrichment | Adds data | CompanyInput | EnrichedCompany |
|
|
| ICP Strategist | Determines ideal customers | CompanyIntelligence | ICPList |
|
|
| Partnership Strategist | Maps partnership types | CompanyIntelligence | PartnershipMap |
|
|
| Channel Strategy | Picks best channel | Company + Compliance | ChannelPlan |
|
|
| Message Generation | Writes Arabic messages | Company + Channel | OutreachMessage |
|
|
| Compliance | Enforces platform rules | Channel + Action | Decision |
|
|
| Campaign Orchestrator | Builds sequences | Company + Messages | CampaignSequence |
|
|
| Negotiation | Handles objections | Reply + Context | NegotiationResponse |
|
|
| CRM & Revenue | Tracks status | Events | StatusUpdate |
|
|
| Learning | Improves system | Results | UpdatedStrategy |
|
|
|
|
## Automation Boundaries
|
|
|
|
| Level | What | Examples |
|
|
|-------|------|---------|
|
|
| Fully Automated | Internal processing | Research, scoring, drafts, CRM tasks, reports |
|
|
| Semi-Automated | Approved channels | Email with opt-out, inbound chatbot, WhatsApp templates |
|
|
| Manual Required | Risky channels | LinkedIn DMs, Instagram DMs, phone calls |
|
|
| Prohibited | Policy violation | LinkedIn scraping, WhatsApp blast, fake accounts |
|
|
|
|
## Safety Rules
|
|
- LinkedIn: NO scraping/bots/automated DMs
|
|
- X: NO unsolicited automated replies/mentions
|
|
- WhatsApp: opt-in only, stop on "إيقاف"
|
|
- Instagram: inbound/official API only
|
|
- TikTok: content + official ads/lead forms only
|