15:52:44 <RRSAgent> RRSAgent has joined #webmachinelearning
15:52:48 <RRSAgent> logging to https://www.w3.org/2026/03/05-webmachinelearning-irc
15:52:48 <Zakim> RRSAgent, make logs Public
15:52:49 <Zakim> please title this meeting ("meeting: ..."), anssik
15:52:50 <anssik> Meeting: WebML CG Teleconference – 5 March 2026
15:52:52 <anssik> Chair: Anssi
15:52:56 <anssik> Agenda: https://github.com/webmachinelearning/meetings/blob/main/telcons/2026-03-05-cg-agenda.md
15:53:01 <anssik> Scribe: Anssi
15:53:05 <anssik> ScribeNick: anssik
15:53:13 <anssik> Present+ Anssi_Kostiainen
15:53:22 <anssik> RRSAgent, draft minutes
15:53:23 <RRSAgent> I have made the request to generate https://www.w3.org/2026/03/05-webmachinelearning-minutes.html anssik
15:54:02 <anssik> gb, this is webmachinelearning/webmcp
15:54:02 <gb> anssik, OK.
15:59:37 <zolkis> zolkis has joined #webmachinelearning
16:00:41 <BenGreenstein> BenGreenstein has joined #webmachinelearning
16:00:43 <anssik> Anssi: to start, we will welcome our latest new participants:
16:00:48 <anssik> ... Omri Belavad and Guilherme Gervasio joining as an individual contributor
16:00:52 <anssik> ... welcome all to the WebML Community Group!
16:00:54 <kush> kush has joined #webmachinelearning
16:00:56 <anssik> ... as a reminder, we'll use IRC-based queue management in this meeting:
16:00:58 <kush> present+
16:01:00 <anssik> -> https://www.w3.org/guide/meetings/zakim.html#speakerqueue
16:01:09 <anssik> Anssi: to suggest agenda topics, use Agenda+ label -- we may discuss newly Agenda+ labeled issues #51 and #130 today too
16:01:18 <gb> https://github.com/webmachinelearning/webmcp/issues/130 -> Issue 130 Tool unregistration design (by domfarolino) [Agenda+]
16:01:18 <gb> https://github.com/webmachinelearning/webmcp/issues/51 -> Issue 51 Define the API for in-page Agents to use a site's declared tools (by khushalsagar) [Agenda+]
16:01:24 <anssik> -> Agenda+ https://github.com/webmachinelearning/webmcp/labels/Agenda+
16:01:30 <anssik> Topic: WebMCP
16:01:36 <anssik> Subtopic: Announcement: Awesome WebMCP
16:01:40 <anssik> Anssi: we launched a new community-curated resource to recognize applicable contributions from the wider WebMCP ecosystem:
16:01:45 <anssik> -> Awesome WebMCP https://github.com/webmachinelearning/awesome-webmcp
16:01:50 <Victor> Victor has joined #webmachinelearning
16:02:18 <brwalder> brwalder has joined #webmachinelearning
16:02:22 <anssik> Subtopic: Tool infra for registerTool/unregisterTool
16:02:27 <anssik> Anssi: PR #113
16:02:28 <gb> https://github.com/webmachinelearning/webmcp/pull/113 -> MERGED Pull Request 113 Add basic tool infra and complete `registerTool()` (by domfarolino) [Agenda+]
16:02:45 <qcomp> qcomp has joined #webmachinelearning
16:02:45 <anssik> ... we addressed all the review comments and landed the PR ahead the meeting, great work
16:02:45 <AlexN> AlexN has joined #webmachinelearning
16:02:45 <anssik> ... incremental improvements will arrive in subsequent PRs
16:02:49 <anssik> ... to that end, I'll propose we move on to the next topic
16:02:53 <anssik> ... any comments?
16:02:58 <Mark_Foltz> Mark_Foltz has joined #webmachinelearning
16:03:05 <Ehsan> Ehsan has joined #webmachinelearning
16:03:09 <anssik> Dominic: this was landed and other work well under way
16:03:10 <anssik> q?
16:03:29 <anssik> q?
16:03:40 <anssik> Subtopic: Declarative explainer
16:03:44 <anssik> Anssi: PR #76
16:03:45 <gb> https://github.com/webmachinelearning/webmcp/pull/76 -> Pull Request 76 Declarative API Explainer (by domfarolino) [Agenda+]
16:03:50 <anssik> ... great review comments, thank you all
16:04:00 <anssik> ... does the group have specific proposals to discuss today?
16:04:10 <domfarolino> domfarolino has joined #webmachinelearning
16:04:15 <anssik> ... are there any objection to approve and merge the baseline explainer with a condition that all open review comments have dedicated issue to ensure we follow them through?
16:04:23 <domfarolino> q+
16:04:29 <anssik> ack domfarolino
16:04:39 <johannhof> johannhof has joined #webmachinelearning
16:05:06 <anssik> Dominic: it seems like all things are resolved exept the last few days worth of comment, JSON-LD discussion
16:05:56 <anssik> ... JSON-LD cross document ergonomics is an open issue for discussion
16:06:03 <anssik> -> https://github.com/webmachinelearning/webmcp/pull/76/changes/b449df1c4214ab376804a31d4c23d135f4b8799b#r2886459866
16:06:04 <gb> https://github.com/webmachinelearning/webmcp/pull/76 -> Pull Request 76 Declarative API Explainer (by domfarolino) [Agenda+]
16:06:14 <anssik> q?
16:06:42 <vasilii> present+ Vasilii_Trofimchuk
16:06:48 <anssik> Alex: I did a prototype and found out these is probably a better primitive
16:06:49 <anssik> q?
16:06:51 <brwalder> q+
16:06:57 <anssik> ack brwalder
16:07:30 <anssik> Brandon: one thought I had, scraping the DOM for JSON-LD and using that as output carries with it an implicit assumption the agent is doing DOM scraping in addition to WebMCP
16:07:38 <kush> q+
16:07:58 <anssik> ... true for in-browser agents possibly, but for other agents don't want to do DOM scraping to get their output, want to use WebMCP only
16:08:24 <anssik> ... having a native WebMCP way to get that state on a new page seems more friendly to agents who don't want to scrape the DOM
16:08:31 <anssik> Alex: extension agents?
16:08:38 <anssik> Brandon: also in-page agents
16:09:10 <anssik> Dominic: that makes sense, I think the trick is for non-built in browser agents, they will have some hack that tells them the context of the DOM is an answer to the previous tool call
16:09:22 <anssik> ... in JS the entire context gets reset after navigation
16:09:43 <anssik> ... JSON-LD only works for built-in agents, and even for that it is a bit hacky
16:09:50 <anssik> ... will file an issue for this
16:09:50 <anssik> q?
16:09:55 <kush> q-
16:10:00 <anssik> q?
16:10:48 <anssik> Anssi: issue #126 is a spin-off from PR #76, interest to discuss this issue today?
16:10:49 <gb> https://github.com/webmachinelearning/webmcp/pull/76 -> Pull Request 76 Declarative API Explainer (by domfarolino) [Agenda+]
16:10:49 <gb> https://github.com/webmachinelearning/webmcp/issues/126 -> Issue 126 Where to fire `toolactivated` and `toolcanceled` events? (by domfarolino) [declarative]
16:11:12 <anssik> Dominic: do we want to fire these on form elements or on the ModelContext object or Window
16:11:16 <chrishtr> chrishtr has joined #webmachinelearning
16:11:33 <anssik> ... to support these events on imperative tools, we cannot get away from this
16:11:33 <brwalder> q+
16:12:05 <anssik> ack brwalder
16:12:30 <anssik> Brandon: Dominic good points on your the most recent comment, if imperative tools need to do anything they can just use the execute function
16:12:49 <anssik> ... putting the event on ModelContext object seems like the best design
16:12:55 <anssik> q?
16:14:12 <domfarolino> s/"live"/"are fired"
16:14:39 <domfarolino> +1
16:14:39 <anssik> RESOLUTION: The toolactivated and toolcanceled events are fired on the ModelContext object. (issue #126)
16:14:39 <gb> https://github.com/webmachinelearning/webmcp/issues/126 -> Issue 126 Where to fire `toolactivated` and `toolcanceled` events? (by domfarolino) [declarative]
16:14:49 <anssik> q?
16:15:10 <anssik> Subtopic: Agent allowlist use cases and requirements
16:15:13 <anssik> Anssi: issue #116
16:15:14 <gb> https://github.com/webmachinelearning/webmcp/issues/116 -> Issue 116 Agent allowlist use cases and requirements (by anssiko) [Agenda+]
16:15:24 <anssik> ... I think the M:N problem illustrated by Yoav's informs this feature design:
16:15:28 <anssik> -> https://github.com/webmachinelearning/webmcp/issues/51#issuecomment-3973968414
16:15:29 <gb> https://github.com/webmachinelearning/webmcp/issues/51 -> Issue 51 Define the API for in-page Agents to use a site's declared tools (by khushalsagar) [Agenda+]
16:15:54 <anssik> Anssi: also Alex shared related learnings from the crypto space, see "Pointer Locks for Agents & Agent whitelisting"
16:15:59 <anssik> -> https://github.com/webmachinelearning/webmcp/issues/43#issuecomment-3478492067
16:16:00 <gb> https://github.com/webmachinelearning/webmcp/issues/43 -> CLOSED Issue 43 Clarifying the scope of the proposal (by 43081j)
16:16:27 <anssik> Anssi: for WebMCP too, we could have a future where we have in-page agents, iframe agents, extension-based agents and in-browser agents all living together on the web platform and talking WebMCP
16:16:47 <anssik> ... and all these agents would in principle know how to operate the tools provided by any website
16:16:51 <anssik> ... based on past learnings, it seems the agent allowlist mechanism should be considered together with an agent negotiation/lock mechanism #118
16:16:52 <gb> https://github.com/webmachinelearning/webmcp/issues/118 -> Issue 118 Agent negotiation (by yoavweiss) [backlog]
16:16:55 <domfarolino> q+
16:16:56 <anssik> ... I'll invite participants to share concrete use cases in this issue
16:17:26 <anssik> ack?
16:17:31 <anssik> ack domfarolino
16:17:37 <AlexN> q+
16:18:01 <anssik> Dominic: if we don't introduce an API that allows to enumerate and list, then monkey-patching will  likey happen
16:18:15 <anssik> ... to avoid this registerTool problem, we can introduce a registry
16:19:15 <anssik> ... two sub-problems, A) do we want agent identifier or allowlist to register tools, only exposed to extension agent, in-browser agent etc.; B) or one registry that is visible to all tools
16:19:36 <anssik> ... second is concurrency problem, with pointer lock style as one possible solution
16:20:06 <anssik> ... blocking out all other agents when one agent is executing maybe too heavy-handed solution
16:20:07 <brwalder> brwalder has joined #webmachinelearning
16:20:30 <anssik> ... the simplest solution would be a global registry and design it so we can add on functionality to filter tools from specific agents
16:20:51 <anssik> ... when you listTools() registered and you can view all global tools and those tools exposed to your agent only
16:21:08 <anssik> ... I think this is a decent path forward
16:21:16 <anssik> ... that might help with concurrency management
16:21:24 <anssik> q?
16:21:59 <anssik> Dominic: I think we're most excited to move forward with global registry of tools that can be extended with per-agent filter mechanism
16:22:15 <anssik> ... pointer lock where platform blocks other agents seems a bit heavy
16:22:16 <anssik> q?
16:22:22 <anssik> ack AlexN
16:22:42 <anssik> Alex: that sums up it pretty well, want to highlight Khus' proposal
16:22:45 <kush> q+
16:23:08 <anssik> ... we will see many different agents interacting
16:23:08 <anssik> q?
16:23:15 <anssik> ack kush
16:24:08 <anssik> Khushal: I was convinced it seems premature to solve, filtering which tool goes to which agent, both the things don't need to be solved in the same agent and right now, we just want to make sure this API is extensible to addess these both things, filtering and ensuring only one agent is running at a time
16:25:10 <anssik> Dominic: personally I'd like to see real concurrency handled by the web developer ultimately, basic global list of tools that extends to per-agent registry
16:25:12 <anssik> q?
16:25:40 <anssik> Anssi: are we clear on the use cases?
16:26:41 <anssik> Dominic: I think the use case would need to clerified when we are to augment the global registry, understand multiple agents stomping each other
16:26:42 <anssik> q?
16:26:57 <anssik> q?
16:27:19 <Kryspin_Ziemski> q+
16:27:22 <AlexN> +1
16:27:23 <kush> +1
16:27:27 <anssik> Dominic: my plan is to spec this and make sure it is extensible for a filtering mechanism
16:28:16 <anssik> Brandon: sounds good
16:28:17 <anssik> q?
16:28:27 <anssik> ack Kryspin_Ziemski
16:29:37 <anssik> Kryspin: we're talking about executing, how do we think about coordination, where there's hierarchy of agents, primary and flowing down from there
16:30:31 <anssik> ... we're talking about tools and concurrency, the tools themselves mutate state because the agent determines it needs to be called, e.g. call checkout function, that may be relevant for in-page agent to know it was called, both aware of the interaction happening
16:31:52 <anssik> ... shared tools are an intention and if in-browser agent calls a function it should bubble down, otherwise they're operating without knowledge of intention
16:32:03 <brwalder> q+
16:32:05 <anssik> ... think shared table, where people are not seeing what others are doing
16:33:13 <anssik> Dominic: if in-browser agent calls a checkout flow, maybe the in-page agent wants to know, it is useful, the same if the user clicks the checkout button in the UI
16:33:29 <anssik> Kryspin: the in-page agent needs to know, to be able to act on the checkout function
16:33:38 <AlexN> +q
16:34:09 <anssik> q?
16:34:27 <anssik> RRSAgent, draft minutes
16:34:28 <RRSAgent> I have made the request to generate https://www.w3.org/2026/03/05-webmachinelearning-minutes.html anssik
16:34:42 <anssik> q?
16:35:06 <anssik> ack brwalder
16:35:14 <Mark_Foltz> q+
16:36:21 <anssik> Brandon: I think some of this is on web developer to solve, we have in-page agent, in-browser agent, to mitigate this is to build in-page agents in a way that's useful to both agents, if there's an action on the page they don't want in-browser agent doing, they shouldn't expose that tool
16:37:08 <anssik> ... need coordination to have the means to delegate between the two
16:37:19 <anssik> q?
16:38:07 <anssik> Kryspin: ModelContext is not registering against an agent, it'd be sense for in-page agent to be given tools, that's limit scope
16:38:52 <Victor> q+
16:39:03 <anssik> ... if no interaction between the two agents, shared prompt mechanism or means to play an assistant role to the other agent, not able to know what the in-browser is tasked to do, from the in-page agent side
16:39:32 <anssik> ... allowlist will solve of this
16:39:32 <anssik> q?
16:39:34 <anssik> ack AlexN
16:40:12 <anssik> Alex: to the points of this sharing state across agents, there's a need for some context that does nto need to be fetched via tool call but written from the app
16:40:22 <anssik> ... solves coordination issue, will open an issue about this
16:40:29 <anssik> q?
16:40:31 <anssik> ack Mark_Foltz
16:41:22 <anssik> Mark: there's several topics here, one thing is to think if there are multiple agents that are concurrently modifying a page, allow feedback to be shared, what changed, to give direction to plan the next tool call
16:43:21 <anssik> Kryspin: there should be a message bus, a messenger pattern, different agents having understanding what other agents are doing, more important than concurrency, in-page agents don't know what in-browser agent is doing
16:43:28 <AlexN> prior art: https://docs.copilotkit.ai/learn/whats-new/v1-50#useagent &  https://developers.cloudflare.com/agents/api-reference/store-and-sync-state/
16:44:09 <anssik> Kryspin: people we do ambitious things, browser agent has more agency than in-page agents, and agents need to coordinate
16:44:22 <anssik> Dominic: return value from the tool dispatched to all agents?
16:44:27 <anssik> ... all agents listening to that?
16:44:50 <anssik> Kryspin: yes, this is the missing piece
16:44:50 <anssik> q?
16:44:56 <anssik> ack Victor
16:45:49 <anssik> Victor: in-page agents, we're trying to solve the problems without good list of use cases for this class of problems
16:46:34 <anssik> Kryspin: I agree with that, we started with tool calling and shared registry of tools
16:46:56 <anssik> ... a separate explainer would be great, help people coming from MCP space to WebMCP space
16:47:37 <anssik> Dominic: we can add to an existing explainer
16:47:50 <anssik> q?
16:48:37 <anssik> Kryspin: I will add to an existing explainer agent coordination section
16:48:55 <anssik> q?
16:49:06 <anssik> Subtopic: Input schema validation ownership
16:49:10 <anssik> Anssi: issue #92
16:49:10 <gb> https://github.com/webmachinelearning/webmcp/issues/92 -> Issue 92 Who owns the validation layer? (by MiguelsPizza) [Agenda+]
16:49:19 <anssik> ... Alex proposes the browser to do the input schema validation and return validation errors in a structured way
16:49:24 <anssik> ... the group seems to agree on that high-level point
16:49:32 <anssik> ... Dominic commented there are three validation layers: meta, input and output validation
16:49:39 <anssik> ... we seem to have an initial agreement that browser should do three-layer validation
16:49:52 <anssik> Anssi: do we want to discuss error behavior on the call today?
16:50:01 <anssik> q?
16:50:38 <anssik> Dominic: meta validation, how strict we want to be?
16:50:59 <anssik> ... JSON schema supports dynamic linking, for example
16:51:16 <anssik> ... what is the normal behaviour, for best alignment with other projects
16:51:16 <anssik> q?
16:52:03 <anssik> Alex: standard schema is what the community has rallied around, it is the best resource
16:52:07 <Kryspin_Ziemski> https://standardschema.dev/
16:52:31 <anssik> q?
16:52:48 <anssik> Dominic: what would be the default look like?
16:52:51 <kush> q+
16:53:15 <anssik> Alex: you can write your own validator
16:53:19 <anssik> q?
16:53:22 <anssik> ack kush
16:53:55 <anssik> Khushal: should ensure interop across all browser agents
16:54:06 <anssik> Dominic: what keywords to support and expand from there
16:54:08 <anssik> q?
16:54:19 <gb> https://github.com/webmachinelearning/webmcp/issues/92 -> Issue 92 Who owns the validation layer? (by MiguelsPizza) [Agenda+]
16:54:55 <anssik> proposed RESOLUTION: Browser is responsible for input schema validation on all three layers: meta, input and output. TODO research JSON schema validation and codify a subset in the spec. (issue #92)
16:55:09 <domfarolino> +1
16:55:21 <kush> +1
16:55:22 <Kryspin_Ziemski> +1
16:55:23 <brwalder> +1
16:55:24 <AlexN> +1
16:55:26 <anssik> RESOLUTION: Browser is responsible for input schema validation on all three layers: meta, input and output. TODO research JSON schema validation and codify a subset in the spec. (issue #92)
16:55:40 <anssik> Subtopic: provideContext overwrite
16:55:44 <anssik> Anssi: issue #101
16:55:44 <gb> https://github.com/webmachinelearning/webmcp/issues/101 -> Issue 101 `navigator.modelContext.provideContext` allows overwriting of previously registered tools in the same environment (by beaufortfrancois) [Agenda+]
16:55:52 <anssik> ... since I put this on the agenda, Brandon shared motivation for the initial provideContext design:
16:55:55 <anssik> -> https://github.com/webmachinelearning/webmcp/issues/101#issuecomment-3987182502
16:55:56 <gb> https://github.com/webmachinelearning/webmcp/issues/101 -> Issue 101 `navigator.modelContext.provideContext` allows overwriting of previously registered tools in the same environment (by beaufortfrancois) [Agenda+]
16:56:05 <anssik> .. and informed by this background and current focus on tools, Brandon proposed to drop provideContext/clearContext from the spec entirely
16:56:09 <anssik> ... and it seems the group agrees
16:56:17 <qcomp> qcomp has joined #webmachinelearning
16:56:25 <anssik> q?
16:56:44 <domfarolino> q+
16:56:51 <anssik> ack domfarolino
16:57:29 <anssik> Dominic: I put up a PR to do this spec change, also doing this in Chrome, I see no controversy doing this change
16:57:31 <anssik> q?
16:58:12 <domfarolino> https://github.com/webmachinelearning/webmcp/pull/132
16:58:12 <gb> https://github.com/webmachinelearning/webmcp/pull/132 -> Pull Request 132 Remove `provideContext()` and `clearContext()` (by domfarolino)
16:58:16 <gb> https://github.com/webmachinelearning/webmcp/pull/132 -> Pull Request 132 Remove `provideContext()` and `clearContext()` (by domfarolino)
16:58:16 <gb> https://github.com/webmachinelearning/webmcp/issues/101 -> Issue 101 `navigator.modelContext.provideContext` allows overwriting of previously registered tools in the same environment (by beaufortfrancois) [Agenda+]
16:58:30 <domfarolino> +1
16:58:35 <kush> +1
16:58:35 <Victor> +1
16:58:35 <AlexN> +1
16:58:35 <anssik> RESOLUTION: Drop provideContext/clearContext from the spec as in PR #132. (issue #101)
16:59:50 <anssik> Subtopic: Define the API for in-page Agents to use a site's declared tools
16:59:54 <anssik> Anssi: issue #51 was proposed to the agenda yesterday
16:59:55 <gb> https://github.com/webmachinelearning/webmcp/issues/51 -> Issue 51 Define the API for in-page Agents to use a site's declared tools (by khushalsagar) [Agenda+]
16:59:56 <AlexN> +1
17:00:07 <anssik> ... this discussion was reignited by Yoav's M:N problem illustration, complexity explosion that exist between "MCP Servers" and "MCP Clients" in context of WebMCP when we use more loosely scoped definitions
17:00:13 <anssik> ... we revisited this in context of agent allowlist discussion:
17:00:16 <anssik> -> https://github.com/webmachinelearning/webmcp/issues/51#issuecomment-3973968414
17:00:17 <gb> https://github.com/webmachinelearning/webmcp/issues/51 -> Issue 51 Define the API for in-page Agents to use a site's declared tools (by khushalsagar) [Agenda+]
17:00:25 <anssik> Anssi: Brandon's proposal is to move listTools and executeTools to navigator.modelContext to allow non-browser agents enumerate and execute tools
17:00:40 <kush> +1
17:00:40 <brwalder> +1
17:00:41 <domfarolino> +1
17:00:47 <qcomp> +1
17:00:50 <Victor> +1
17:00:52 <AlexN> +1
17:00:53 <anssik> RESOLUTION: Move listTools and executeTools to navigator.modelContext to allow non-browser agents enumerate and execute tools. (issue #51)
17:01:07 <anssik> RRSAgent, draft minutes
17:01:08 <RRSAgent> I have made the request to generate https://www.w3.org/2026/03/05-webmachinelearning-minutes.html anssik
17:01:45 <reillyg> Great meeting!
17:01:54 <AlexN> Thanks anssik!
17:02:19 <anssik> RRSAgent, draft minutes
17:02:21 <RRSAgent> I have made the request to generate https://www.w3.org/2026/03/05-webmachinelearning-minutes.html anssik
17:02:39 <anssik> Present+ Mark_Foltz
17:02:43 <anssik> Present+ Dominic_Farolino
17:02:52 <anssik> Present+ Kryspin_Ziemski
17:03:06 <anssik> Present+ Alex_Nahas
17:03:12 <anssik> Present+ Ben_Greenstein
17:03:18 <anssik> Present+ Brandon_Walderman
17:03:36 <anssik> Present+ Chris_Harrelson
17:03:42 <anssik> Present+ Ehsan_Toreini
17:04:00 <anssik> Present+ Iris_Johanson
17:04:08 <anssik> Present+ Khushal_Sagar
17:04:17 <anssik> Present+ Reilly_Grant
17:04:29 <anssik> Present+ Sathish
17:04:41 <anssik> Present+ Johann_Hofmann
17:05:00 <anssik> Present+ Tania_Millan
17:05:04 <anssik> Present+ Victor_Huang
17:05:19 <anssik> Present+ Zoltan_Kis
17:05:24 <anssik> RRSAgent, draft minutes
17:05:26 <RRSAgent> I have made the request to generate https://www.w3.org/2026/03/05-webmachinelearning-minutes.html anssik
17:08:00 <anssik> s/these is/there is
17:09:30 <anssik> s/your the most/your most
17:10:25 <anssik> s/informs/comment informs
17:11:20 <anssik> s/likey/likely
17:11:56 <anssik> s/maybe too/may be too
17:13:14 <anssik> s/clerified/clarified
17:15:04 <anssik> s/be sense/make sense
17:15:27 <anssik> s/that's limit/that'd limit
17:15:53 <anssik> s/solve of this/solve this
17:16:11 <anssik> s/nto/not
17:17:09 <anssik> s/people we/people will
17:20:54 <anssik> RRSAgent, draft minutes
17:20:55 <RRSAgent> I have made the request to generate https://www.w3.org/2026/03/05-webmachinelearning-minutes.html anssik
19:33:27 <Zakim> Zakim has left #webmachinelearning