Skip to content

Agent Corps and Operational Roles

SocioProphet uses a corps of narrowly-scoped operational roles rather than one unconstrained “AI that does everything.”

This is a public-safe summary of those roles.

1. Public role model

The public-facing roles include:

  • observer and profiler
  • egress and policy steward
  • safeguard and enforcement role
  • provenance and ledger role
  • simulation and validation role
  • evaluation and diffusion role
  • orchestration role
  • privacy and DLP role
  • bridge and diffusion controller

2. Why role separation matters

Role separation reduces risk.

It prevents:

  • one component from holding every privilege
  • hidden authority chains
  • unsafe coupling of detection, enforcement, and promotion
  • unreviewable automation drift

3. What the corps does publicly

Publicly, we document the corps in terms of:

  • mission
  • scope
  • bounded capabilities
  • proof outputs
  • safety gates
  • operator handoff model

4. What we do not publish publicly

We do not publish the full internal operator kit.

We do not publish:

  • high-risk playbooks
  • sensitive enforcement internals
  • adversary-emulation mechanics
  • private thresholds
  • exploit or persistence methods
  • misuse-enabling tactical details

5. Relation to the platform

The corps belongs inside the governed operational model described in:

6. Public takeaway

The point is not to build an army of opaque bots. The point is to create a disciplined, bounded, evidence-producing operational stack with clear roles, proofs, and review.