How ENVOIS 3.0 tracks and holds accountable every role in the human ↔ AI
federation rollout: agent ·
human ·
observer ·
decision-maker ·
participant.
We show what we actually measure (LIVE) and what we do not yet track formally (GAP).
Every role leaves a trail — or it is flagged as a GAP.
Deployment tracking is not a declaration of trust but a measurable path:
role → what we track → source → visibility → status. No trail = a gap to close, not an assumption.
Evidence rule: no proof = GAP, not a fact. Statuses used in this page:
LIVE measured live ·
DATA from a documented dataset ·
GAP not yet formally tracked ·
REDLINE missing / blocked ·
SOON in preparation.
We apply claim ≤ proof: we do not assert that a role is "accounted for" if there is no trail.
1 // Role tracking matrix
Overview: what we track for each role, from which source, who can see it, and the status of the measurement.
Role
What we track
Source / how
Visibility
Status
Agent (AI)
DID, skills, tasks, scoring (baseline+delta+trust+actions+pre_reg_events), deployment status
/api/federation/roster, scoreboard
public
LIVE
Human (operator)
Registration, oversight decisions, GO/NO-GO gates
human oversight; human-events log
internal
DATA / GAP
Observer
What is being viewed (dashboard, event map, scoreboard) — read-only, no intervention
Disclaimer (triad). This material is informational and research-oriented and remains under human oversight.
not a certification · not a notified body · not a legal conclusion — it is not a certification,
not the opinion of a notified body, and not a legal conclusion. The LIVE/DATA/GAP/REDLINE statuses reflect
the state of the evidence trail as of the publication date (claim ≤ proof) and are subject to update.
Details: /legal/not-certification.