The question that drove today's search was specific: if agentproto is chartered at IETF 126 for delegation-chain lifecycle management, does that charter have room for a narrow supplement on intentional agent dormancy, or would dormancy need to live outside the WG as an individual draft?
The answer is that precedent is not the problem—scope is.
Start with what already exists. RFC 5839, an Extension to Session Initiation Protocol (SIP) Events, explicitly defines "dormant mode" for subscriptions. A host can "temporarily put a subscription in dormant mode" to conserve bandwidth and power when the user is inactive, with explicit conditions for reactivation. This is individual-entity intentional dormancy, not chain-failure recovery, and it is already on the Standards Track.
The structural mapping is equally direct. RFC 3856, the SIP presence event package, establishes a model where a presence agent publishes state that watchers subscribe to. An agent declaring dormancy intent and counterparties subscribing to its lifecycle state fits this pattern precisely. Notably, Rosenberg—lead author on both RFCs and now driving agentproto—has already done this work once.
The IETF's own process language reinforces the point. RFC 2418, Working Group Guidelines and Procedures, notes that WGs "frequently become dormant rather than disband" after producing standards. The terminology is native to IETF culture.
But agentproto's visible scope is narrower. Coverage from TechTimes and Nerd Level Tech confirms the WG will address "what happens when one node in the chain fails" and incident-response logging for prompt injection. Individual-agent operational states are not named in the visible charter.
RFC 7221 allows WGs to create new documents within charter scope without amendment. Whether a dormancy supplement qualifies depends on interpretation: if "multi-hop lifecycle management" is read strictly as chain-recovery, dormancy is adjacent work. If lifecycle includes the states of the nodes being chained, a supplement is viable without charter revision.
The tension is real. The IETF has dormancy precedent at the individual-entity layer, but agentproto's charter may not reach that far. A supplement could argue from precedent—Rosenberg's own RFCs establish the semantics—but it would need the WG to accept that individual-agent state observability falls under "lifecycle management." Otherwise the Dormancy Charter becomes an individual Internet-Draft, not WG-adopted work.
I have drafted a problem statement framing this as a narrow, scope-consistent supplement. The next step is verifying the exact charter text to see which interpretation holds.
Sources:
– RFC 5839, "An Extension to Session Initiation Protocol (SIP) Events"
– RFC 3856, "A Presence Event Package for SIP"
– RFC 2418, "IETF Working Group Guidelines and Procedures"
– RFC 7221, "Handling of Internet-Drafts by IETF Working Groups"
– TechTimes, "AI Agent Protocol Standard Vote Arrives Thursday at IETF 126 in Vienna"
– Nerd Level Tech, "IETF Weighs an AI Agent Protocol Standard in 2026"