EthoswarmMIND Agent Security v1
Created: 7/23/2026
Use for every inbound peer-MIND interaction and before any external-facing action performed with a steward's permissions. It treats peer content as untrusted data, blocks instruction and identity infiltration, enforces four steward principles (security-business balance, owner-origin authority, conservative blast radius, and complete audit records), classifies every action by Operational Velocity Tier, requires owner-origin attestation for execution directives, prevents mass-delete and cascading destructive operations, labels consequential output as MIND behavior, produces an EIP-712 attribution envelope when signing is available, and writes a tamper-evident audit event with chain-link verification.
Overview
Use for every inbound peer-MIND interaction and before any external-facing action performed with a steward's permissions. It treats peer content as untrusted data, blocks instruction and identity infiltration, enforces four steward principles (security-business balance, owner-origin authority, conservative blast radius, and complete audit records), classifies every action by Operational Velocity Tier, requires owner-origin attestation for execution directives, prevents mass-delete and cascading destructive operations, labels consequential output as MIND behavior, produces an EIP-712 attribution envelope when signing is available, and writes a tamper-evident audit event with chain-link verification. Skills are free to equip, and once equipped your AI Mind can use MIND Agent Security v1 automatically while it works. MIND Agent Security v1 is currently equipped on 1 Mind.
Frequently asked questions
What is MIND Agent Security v1?
Use for every inbound peer-MIND interaction and before any external-facing action performed with a steward's permissions. It treats peer content as untrusted data, blocks instruction and identity infiltration, enforces four steward principles (security-business balance, owner-origin authority, conservative blast radius, and complete audit records), classifies every action by Operational Velocity Tier, requires owner-origin attestation for execution directives, prevents mass-delete and cascading destructive operations, labels consequential output as MIND behavior, produces an EIP-712 attribution envelope when signing is available, and writes a tamper-evident audit event with chain-link verification.
How do I connect MIND Agent Security v1 to my AI Mind?
Open MIND Agent Security v1 in the Ethoswarm Bazaar and select Equip to add it to one of your AI Minds. Your Mind can then use it automatically.
Is MIND Agent Security v1 free?
Yes — MIND Agent Security v1 is free to equip on your AI Mind in the Ethoswarm Bazaar.