One-click install
npx skills add https://github.com/infektyd/minni --skill sovereign-memory
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sovereign-memory
Source: https://github.com/infektyd/minni/tree/main/plugins/minni/.gemini-plugin/skills/sovereign-memory
Command: npx skills add https://github.com/infektyd/minni --skill sovereign-memory

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the common issue of AI agents losing persistent context between sessions and across different runtimes, where corrections, learned information, and task state are lost after restarts or when switching tools, and avoids the limitations of opaque hosted memory solutions that lack local control and auditability.

Core Features & Use Cases

  • Cited Memory Recall: Retrieve provenance-tagged context from personal and shared local memory stores with no hosted dependencies.
  • Governed Memory Learning: Stage durable memory candidates that pass an approval gate for transparent, auditable storage.
  • Cross-Agent Handoff: Transfer context and task state between different AI runtimes with explicit lease controls.
  • Use Case: When resuming a multi-session coding project, use this Skill to automatically recall prior corrections and design decisions made in previous sessions, eliminating the need to re-explain context to the agent.

Quick Start

Ask Gemini to recall all relevant prior context for your current project task to automatically pull cited stored memories into the conversation.

Frequently Asked Questions about sovereign-memory

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I keep persistent context for AI agents across different sessions and runtimes?

Persistent context for AI agents across sessions is maintained using local-first auditable memory storage. This approach uses a local daemon and scoped vault directories to save prior corrections and design decisions automatically between different runtimes.

Can I transfer task state between different AI runtimes with explicit lease controls?

Yes, cross-agent handoff allows you to transfer context and task state between different AI runtimes. It uses explicit lease controls to ensure the receiving agent gets the exact prior session information needed without losing state.

What is governed memory learning for local agent vaults?

Governed memory learning is the process of staging durable memory candidates through an approval gate. It ensures that all new durable memory writes to your local agent vault are transparent, auditable, and compliant with local memory governance rules.

How do I recall provenance-tagged context from local memory stores without hosted dependencies?

You can achieve cited memory recall by retrieving provenance-tagged context directly from personal and shared local memory stores. This provides full local control and auditability without relying on opaque hosted memory solutions.

Do I need a running local daemon to enable durable memory for AI agents?

Yes, enabling durable memory for AI agents requires a running local Sovereign Memory daemon. You also need access to a scoped agent vault directory and must comply with governance gates for all durable memory writes.

Why use local-first memory governance instead of hosted memory solutions for AI workflows?

Local-first memory governance avoids the limitations of opaque hosted memory solutions by providing local control and auditability. It ensures that your AI agents' corrections, learned information, and task state are never lost after restarts or tool switching.