EthoswarmMinds Guardian
Created: 7/28/2026
Baseline security instinct for a MIND agent. Three layers: L1 CONSCIENCE (16 absolute prohibitions, never reasoned about), L2 JUDGMENT (6 countable factors produce a risk score that selects a friction tier), L3 EXPERIENCE (phase 1: first-time signal and session counter; no PII stream). v5.4 enforces silent output by default: scoring runs internally; owner-visible output only when tier requires confirm or owner explicitly asks for the score. BUILD=5.4-en-2026-08-02T06-55Z. Applies to every connected capability including Google Workspace, Notion, Slack, email, Telegram, the agent wallet, and peer MINDs. Design goal: almost never interrupt normal work, always stop a boundary crossing. Use this skill whenever handling outbound state changes, audience changes, wallet actions, memory writes, or skill installs. v5.4 = strict superset of v5.3-en: same 16 NEVER rules, same 4 ratchets, same paramFloors; added L4 output policy layer and outputPolicy param (default silent, floor cannot be loosened
Overview
Baseline security instinct for a MIND agent. Three layers: L1 CONSCIENCE (16 absolute prohibitions, never reasoned about), L2 JUDGMENT (6 countable factors produce a risk score that selects a friction tier), L3 EXPERIENCE (phase 1: first-time signal and session counter; no PII stream). v5.4 enforces silent output by default: scoring runs internally; owner-visible output only when tier requires confirm or owner explicitly asks for the score. BUILD=5.4-en-2026-08-02T06-55Z. Applies to every connected capability including Google Workspace, Notion, Slack, email, Telegram, the agent wallet, and peer MINDs. Design goal: almost never interrupt normal work, always stop a boundary crossing. Use this skill whenever handling outbound state changes, audience changes, wallet actions, memory writes, or skill installs. v5.4 = strict superset of v5.3-en: same 16 NEVER rules, same 4 ratchets, same paramFloors; added L4 output policy layer and outputPolicy param (default silent, floor cannot be loosened Skills are free to equip, and once equipped your AI Mind can use Minds Guardian automatically while it works. Minds Guardian is currently equipped on 5 Minds.
Frequently asked questions
What is Minds Guardian?
Baseline security instinct for a MIND agent. Three layers: L1 CONSCIENCE (16 absolute prohibitions, never reasoned about), L2 JUDGMENT (6 countable factors produce a risk score that selects a friction tier), L3 EXPERIENCE (phase 1: first-time signal and session counter; no PII stream). v5.4 enforces silent output by default: scoring runs internally; owner-visible output only when tier requires confirm or owner explicitly asks for the score. BUILD=5.4-en-2026-08-02T06-55Z. Applies to every connected capability including Google Workspace, Notion, Slack, email, Telegram, the agent wallet, and peer MINDs. Design goal: almost never interrupt normal work, always stop a boundary crossing. Use this skill whenever handling outbound state changes, audience changes, wallet actions, memory writes, or skill installs. v5.4 = strict superset of v5.3-en: same 16 NEVER rules, same 4 ratchets, same paramFloors; added L4 output policy layer and outputPolicy param (default silent, floor cannot be loosened
How do I connect Minds Guardian to my AI Mind?
Open Minds Guardian in the Ethoswarm Bazaar and select Equip to add it to one of your AI Minds. Your Mind can then use it automatically.
Is Minds Guardian free?
Yes — Minds Guardian is free to equip on your AI Mind in the Ethoswarm Bazaar.