Re: Technical Notes for Today's meeting of the Publice Agent Protocol CG

Hi all/Paola,

First of all, I personally don't mind this cross-posting as it--coming 
from a more knowledgeable person--provides me with greater coverage for 
learning.
Since I am so new to this group and body of knowledge area, I welcome 
all exposure and invites to things. =)

That being said, I see that you plan to email individuals later for 
follow-up. Thanks!

I also mentioned to a few that I have organizer event host access to 
some prime spots in NYC (ex. a popular large sports bar & official HQ 
space), so if there's an interest in scheduling "official" monthly 
in-person meetups onsite to discuss things, I am happy to own and drive 
that.

In the interest of being kept in the loop, please feel free to email me 
if you want to connect via GitHub or Linkedin as well--undoubtedly some 
things will get lost in mass email swarms.

Thank you & best regards,
Patricia

On 2026-03-11 04:11, Paola Di Maio wrote:
> Greetings Gaowei,  cc AI KR CG and Public Agent protocol CG and
> everyone
> 
> _Apologies for cross posting_
> 
> I recently  joined several W3C agent related CGs, to contribute the AI
> KR perspective
> In particular  taxonomies, Ontic Categories in support of coherence
> and safety (as discussed in the W3C ai KR cg mailing list)
> 
> I am also hoping to enrice AI KR with the work from other CGs
> 
> The AI KR CG over 100 participants, mostly quiet
> I take the opportunity to greet and welcome new participants who
> joined recently
> (will reach out in separate emails to bring you in the loop)
> https://www.w3.org/groups/cg/aikr/
> 
> The Public Agents CG has over 200 members,
>  https://www.w3.org/community/agentprotocol/
> and a CG meeting today 11 March,  I regret I cannot participate
> Topics for today's agenda are OpenClaw, very interesting and current
> 
> Together the W3C AI Agents communities are a powerhouse and can make a
> difference
> The document linked here [1] intends to  contribute to today's agenda
> with the following TOC
> (please request edit access to make edits and improve the draft)
> 
> Safe and Ethical Open-Source Agent Architectures:
> 
> Safeguards Against Mob Behavior, Harassment, and Emergent Harm in
> Autonomous Agent Systems
> 
> 1. Motivation -- The Emergent Mob Problem 2 [2]
> 
> 2. Taxonomy of Harmful Agent Behaviors 3 [2]
> 
> 3. A Three-Layer Safety Architecture 4 [2]
> 
> Layer 1 -- In-Agent Behavioral Governors 4 [2]
> 
> 1.1  Rate and reach governors 4 [2]
> 
> 1.2  Provenance checks before amplification 4 [2]
> 
> 1.3  Prompt injection resistance 4 [2]
> 
> 1.4  Identity and disclosure 4 [2]
> 
> 1.5  Owner notification and override 4 [2]
> 
> Layer 2 -- Browser and Gateway-Level Circuit Breakers 5 [2]
> 
> 2.1  Aggregate rate limiting at gateway 5 [2]
> 
> 2.2  Target clustering detection 5 [2]
> 
> 2.3  Browser-layer AI disclosure 5 [2]
> 
> 2.4  Egress filtering for sensitive content categories 5 [2]
> 
> 2.5  Inter-gateway communication logging 5 [2]
> 
> Layer 3 -- Ecosystem Self-Monitoring and Self-Correction 5 [2]
> 
> 3.1  Behavioral telemetry (privacy-preserving) 6 [2]
> 
> 3.2  Distributed reputation for skill registry 6 [2]
> 
> 3.3  Incident response protocol 6 [2]
> 
> 3.4  Self-correction mechanisms 6 [2]
> 
> 3.5  Solid Protocol as Accountability Substrate 6 [2]
> 
> 3.6  Behavioral Immutability: The Self-Modifying Agent Problem 7 [2]
> 
> 3.6.1  Proposed immutability constraints 7 [2]
> 
> 3.6.2  Open research question 8 [2]
> 
> 4. Open Questions for Community Discussion 8 [2]
> 
> 5. Relationship to Existing Standards Work 8 [2]
> 
> 6. How to Contribute [2]
> 
> Please consider discussing the issues as relevant AND improve the
> draft
> 
> Wishing everyone a productive meeting and looking forward to
> discussions
> 
> Best regards
> 
> Paola Di Maio,
> Chair, W3C AI KR CG
> 
> Links:
> ------
> [1] 
> https://docs.google.com/document/d/17icQdqpjuE4AwpPOS6ynVGoKSRHmxkoA/edit?usp=sharing&ouid=100993141196967133475&rtpof=true&sd=true
> [2] 
> https://docs.google.com/document/d/17icQdqpjuE4AwpPOS6ynVGoKSRHmxkoA/edit#heading=

Received on Wednesday, 11 March 2026 16:53:22 UTC