Agent OS v2 Role Architecture

Agent OS v2 Role Architecture An architecture diagram generated by Archify. Role registry · 19 role archetypes · Agent OS v2 repo Role registry 19 role archetypes Core roles · 6 core roles: atlas, business-owner, qa-verifier, research-analyst, knowledge-steward, platform-reliability · Agent OS v2 repo · required Core roles 6 core roles required Department roles · 9 department archetypes · Agent OS v2 repo Department roles 9 department archetypes Specialist roles · 4 specialist archetypes · Agent OS v2 repo Specialist roles 4 specialist archetypes CONSTITUTION.json · authority + capability contract · Agent OS v2 repo CONSTITUTION.json authority + capability contract SOUL.md · judgment + temperament · Agent OS v2 repo SOUL.md judgment + temperament AGENTS.md · shared non-negotiable rules · Agent OS v2 repo AGENTS.md shared non-negotiable rules Business tenants · Brand75 and future packs · Architecture component Business tenants Brand75 and future packs Tenant packs · agent-assignments.json · Agent OS v2 repo · defined-not-activated Tenant packs agent-assignments.json defined-not-activated Activation gates · identity + tools + evals gates · Architecture component Activation gates identity + tools + evals gates Runtime · compose + invoke ephemeral · Agent OS v2 repo Runtime compose + invoke ephemeral core department specialist paired contract paired guidance paired guidance authority envelope judgment every invocation actor assignments defined-not-activated allow invocation Agent OS v2 repo Legend Backend Database Cloud Security External

How the agent team is organized

Agent OS v2 is the role architecture behind Chris's agent teams. It defines what each agent is allowed to do, how it is evaluated, and how roles compose into a runtime that does real work.

The structure:

This is the governance layer that lets an AI team run without a human watching every step.

Brand75's Agent OS framework for clients →

Role model

  • • Each role is a CONSTITUTION.json plus SOUL.md pair
  • • Roles declare capabilities, not installed skills

Activation

  • • Defining a role does not activate it
  • • Brand75 pack status is defined-not-activated
  • • Activation needs actor, scope, tools, authority, evals, and a kill switch

Protected

  • • Roles cannot change their own constitution, SOUL, tool grants, or model policy
  • • Those are version-controlled product changes