Frequently Asked Questions
Everything you need to know about The Sovereignty Protocol
The Basics
Think of it as a governance layer for AI agents β the rules, memory, skills, and workflows that keep them focused, reliable, and self-improving.
At its core it is a meta-cognitive operating system: Laws prevent drift, Skills define what agents can do, Roles assign responsibilities, and Flows orchestrate multi-step work β all enforced persistently across every session.
On top of that governance layer sits a full platform: a hosted AI workforce, scheduled autonomous tasks, smart memory, web crawling, mail, and more β all running under the same set of laws.
Smart Memory
Smart Memory gives your agents persistent, structured recall across every session β not just a conversation dump.
- Disk-first storage with bidirectional sync to PocketBase β survives restarts
- SHA-256 deduplication: updates existing memories instead of creating duplicates
- Automatic keyword extraction via rule-based analysis + LLM enrichment
- Semantic + fuzzy search with stemming so agents find relevant context even with different phrasing
- IntentMemory: say "save this memoryβ¦" and it routes via MCP directly
- User-scoped vs global context β private memories stay private
- Enrichment traces visible in the UI for full transparency
AI Workforce
The workforce is a team of specialised agents, each with a distinct role and set of skills:
Librarian
Autonomous research β sources, enriches, and writes feature briefs
Spectre
Web crawler β crawls seedlists, extracts signals, feeds news & intel
Documenter
Keeps docs in sync β screenshots, API specs, changelogs
Assessor
Self-assessment β audits agent output, writes governance reports
Linter
Code quality β flags drift, deprecated patterns, broken rules
Enricher
Data enrichment β fills gaps in records, news, and cascades
Hunter
News & CVE intelligence β tracks threats and competitor signals
Cipher
Security β monitors access, detects intrusions, manages API keys
Autonomous Tasks & Cascades
Cascades are multi-step automated workflows. You define a sequence of agent actions β research, enrich, document, review β and the system executes them in order, on a schedule or triggered by an event.
Each step is logged, inspectable, and can be paused or retried from the Mission Control dashboard. Failed steps don't kill the whole cascade β agents attempt self-repair before flagging for review.
Self-Hosting vs Platform
The framework is the downloadable governance package β Laws, Skills, Roles, Flows, Templates, and GEMINI.md. You drop it into any project and your IDE's AI agent immediately operates under its rules. Free, open, model-agnostic.
The platform is thesovereigntyprotocol.com β a fully hosted environment where we run the AI workforce, memory, scheduler, cascades, Spectre crawler, mail relay, and more on your behalf. No infrastructure to manage.
Pricing & Licensing
| Tier | Price | Includes |
|---|---|---|
| Trial | Free | Full platform access for 31 days |
| Free | $0/mo | Framework + core platform (Laws, Roles, basic Skills) |
| Pro | $9/mo | Spectre, MCP tools, advanced Skills, AI Mail, full workforce |
| Enterprise | Custom | Pro + priority support, self-host options, custom integrations |
Ready to get started?