Full Workshop: Better Auth — Paola Estefania, Better Auth

AI Engineer · 40 min · 522 sentences · from YouTube's caption track

Each timecode opens YouTube at the start of that sentence. Line anchors (#s42) are the cue ids in the WebVTT, and every line carries its start and end seconds. All transcripts has every talk, and the whole corpus as one file.

1. 00:12Well, hello everyone.
2. 00:14Can you hear me all?
3. 00:15Okay.
4. 00:15Yes.
5. 00:16Perfect.
6. 00:17Well, welcome to our talk, our workshop actually.
7. 00:22Um, I know maybe you were expecting uh Burgett also.
8. 00:26he couldn't come but we make this workshop together for you.
9. 00:29Okay.
10. 00:30So as you see you have there a QR code you can go ahead and go there is um our page for Asian health protocol so that you have it and I don't have to spell it for you.
11. 00:42Um, okay.
12. 00:43So, my whole idea today is instead of like me telling you stuff and you do this and show you how to do stuff, I want us to take one hour
13. 00:54at least to think about Asian security, Asians because sometimes we are excited about like this new AI era, right?
14. 01:02But we never think about or maybe yes, but not as much as I wished.
15. 01:09What happens when you for example say hey agents what are my maybe my schedule for tomorrow what are send an email have you ever thought of this how many of you like raise your hand like how many of you use AI agents every day
16. 01:27okay higher don't be shy okay great how many of you give them access to your Gmail to your calendar to your personal accounts how many of you oh so much less Okay.
17. 01:39The ones that do like what do you do like you connect an MCP server?
18. 01:43Okay.
19. 01:43Yes.
20. 01:44And they have another one.
21. 01:45Good.
22. 01:46Maybe you connect like your personal token to like endpoint.
23. 01:52Okay.
24. 01:52So what happened when we do this?
25. 01:56We're actually doing like the agent is acting on behalf of us but pretending to be us.
26. 02:03So do you think that's a good idea?
27. 02:07Yes or no?
28. 02:08Who thinks it's a good idea?
29. 02:12Not a great idea.
30. 02:13Why?
31. 02:20Exactly.
32. 02:20And why we didn't thought about this like before.
33. 02:25Okay.
34. 02:25So it doesn't matter.
35. 02:26Here we are.
36. 02:27We are thinking about it.
37. 02:28I always remember uh the time like when internet came like I'm almost 40 so I remember and everybody was so happy about it.
38. 02:36But then started to question like okay everything the whole data is public it's like we are in the same moment today right like we are really excited about AI and then after we start to think about security
39. 02:49but it's good so why I put there hire your agent the analogy I want you to think about let's say you have a company or a part of a company and you hire someone
40. 03:03do you give them you need them to have access two things of the company, right?
41. 03:08So, what you do with these people when you showing or when you show, what did they give to you?
42. 03:13You have a your own email, right?
43. 03:15Your own access, your own credentials.
44. 03:18So, if you think about it in Asian world, what are we doing without agents?
45. 03:25We are it's like we are doing like the CEO credentials, right?
46. 03:27You never say like, okay, this is the credentials of CEO, go and read the email.
47. 03:31No, because everything anything could happen.
48. 03:34This is the same.
49. 03:35The idea will be hire your agent in a sense of give them your agents authority instead of your credentials.
50. 03:43Do you think it's a good idea?
51. 03:45How?
52. 03:45Yes or no?
53. 03:47[laughter]
54. 03:48Okay.
55. 03:48My idea is like I know sometimes um we could be shy but I want I think it would be nice like we can interact a bit because it's a moment
56. 03:57that thanks to a engineer and all of us being here all of your time we can get together and think about something.
57. 04:04I think it's not common, right?
58. 04:05So, make 100% off of it.
59. 04:08Um, okay, great.
60. 04:09So, did you all got the page there, the QR codes?
61. 04:13Yeah.
62. 04:14Great.
63. 04:16So, remember like my idea of saying, "Okay, agent read my emails, right?"
64. 04:24Have you ever thought like how the agent gets to there?
65. 04:27How it gets to your email?
66. 04:30First of all, how do Asians know what they can do?
67. 04:39Huh?
68. 04:46Okay, great.
69. 04:46So, you you do it by hand, right?
70. 04:48It's not that something automatic.
71. 04:53Yes.
72. 04:54Exactly.
73. 04:56But
74. 04:58Exactly.
75. 04:58But some huh
76. 05:01exactly context window also too.
77. 05:03But you always somehow connected something to the Asian that okay this is what you can do right.
78. 05:08So what if we have like an ideal world or that can go to maybe a directory and say oh this is the whole thing I can do and this is all the things I can execute.
79. 05:19That would be a smartest idea right?
80. 05:22Have you ever used a phone book?
81. 05:25Yeah right.
82. 05:26this is safe like who I can call so I go and look in the phone book and then I have how to call it I have the number they will be the same
83. 05:33what what I want to do and how do I call it so that's is our first problem discovery how agents discover what they can do and the idea will be it will be an automatic
84. 05:44way not just us all the time telling how and connecting and stuff right then also there's another part it would be nice to tell the agents or like grant Asians
85. 06:00what they can do and we're not but really specific the ones that connected the Gmail accounts maybe they all gave them read access right just read so we can I don't know don't send or like not remove anything so it's more safe
86. 06:15okay but what if you have so many important uh information there that then if the agent something gets um wrong and somebody else can read your data it's good either.
87. 06:25It's not safe in any way.
88. 06:28So what if we can get all these what the agent can do all these tools as he mentioned and you can grant access to every tool like this yes this no this
89. 06:40reading is okay that would be a better idea right so we think about authorization to that agent not like to act behalf of me and have everything I can do then the agent can do it's like going back to the higher
90. 06:56example that a person like I just hired has all the same access that the CEO or maybe you hire have an assistant uh or a friend that do something for you like in your real life
91. 07:08you will never give your access to all your things right and if you do you change them after so this is the same thing and also okay we have an ideal world when agents can discover what they can do
92. 07:22then we can tell them what they can do or not but how do we trace them down.
93. 07:29How do we trace today what the apps do on our behalf when you authorize them?
94. 07:37Huh?
95. 07:39Exactly.
96. 07:40Audi logs.
97. 07:41So perfect audification, right?
98. 07:50Which agent did what on behalf of which user.
99. 07:54In order to do that, that comes to our main thing.
100. 07:58Give the agent's identity.
101. 08:01If we have if every agent has identity, we can trace them down.
102. 08:05We can know what agent did what when in behalf of which user.
103. 08:09It seems like it's so much so much better, right?
104. 08:13And also what else you can do when you can trace someone regarding to the whole hiring thing, you can fire them if they do something wrong.
105. 08:23You can revoke access in the same way the agent will be revoked if you know who they are.
106. 08:29If something were wrong with today's things, you have to disconnect the whole thing, right?
107. 08:33Yes or no?
108. 08:35So imagine have Oh, this same up.
109. 08:38Sorry for the word.
110. 08:39You can't remove it.
111. 08:40That's it.
112. 08:41So we have tracibility.
113. 08:42We have the whole thing.
114. 08:47Okay.
115. 08:47So as it come out as a sum up we have the discovery issue where the agent can go find the service and read the capabilities authorization what this agent can do can do scope down and
116. 09:01who is this agent are you on board with this do you think it's a good idea you have any questions like champion okay great
117. 09:12[snorts]
118. 09:12have you ever heard of a capabilities raise your hand Yes.
119. 09:18Okay.
120. 09:18Do you all know what a capability is?
121. 09:21Who doesn't know?
122. 09:24Okay.
123. 09:24Almost the whole room.
124. 09:25Okay.
125. 09:26So the idea of using capabilities is instead of using scopes, you know, like scope is like a big thing.
126. 09:34For example, we have the read scope.
127. 09:37So it what what does mean it's like nothing so specific.
128. 09:41The other capability is what can do, but it's more cap down.
129. 09:46So I can actually determine what an action agent can do.
130. 09:51So the idea is going back to the example I put in the beginning.
131. 09:58Let's say you talk to the AI in the chat and you say, "Okay, give me my emails from the last week."
132. 10:06So the agent has an intent to do something, right?
133. 10:10It has to be a way that that intent has to be mapped to a tool or maybe several tools.
134. 10:18So that's kind of the idea.
135. 10:19The tools will be like capabilities and an intent will be matched to that.
136. 10:23So that's what we're solving with discoverability.
137. 10:27If you ever read the the protocol that we are proposing or maybe read it later, you will see um our idea of a well-known Asian configuration endpoint.
138. 10:40If you know about a little bit about identity, you remember OIDC, right?
139. 10:44connect these are the same well known and the idea is like to a nation have like a public place to know how they can interact with a service like everything the kind of encryption
140. 10:55uh etc. So it will be something like this and also a place to list capabilities and stuff.
141. 11:06But what happened for example in a world like today that we still I hope I hope in the future we adopt this but in the meantime all the services they're still working just with O.
142. 11:16So how do we do it in the meantime?
143. 11:21What do we need to agents to be able to know which what they can use from each service?
144. 11:27Let's say Gmail.
145. 11:32We can build something like a directory.
146. 11:35So the idea is um to have something I'm going to see it here.
147. 11:43So let's say most of the you're familiar for open API open API spec.
148. 11:50Yes.
149. 11:51Okay.
150. 11:52So most of the common services we use they all implement open API.
151. 11:58So we have like a kind of a place to we as a developers or engineers we can say okay what this service can do.
152. 12:05So the idea is to use that and convert the two capabilities.
153. 12:10So at the beginning until we every service is speak.
154. 12:14So let's say for example do you guys use I don't know tell me a service you use ah
155. 12:24open AI okay Gmail we have it there but let's say for example let's see if it works open AI open APIJSON and sometimes they have public they have the publics
156. 12:40but we need like a JSON Uh, I don't see it.
157. 12:47API opening.
158. 12:49Okay, I do know.
159. 12:51I think it's this one.
160. 12:53No. Well, you know what?
161. 12:56That were wrong.
162. 12:58Notion.
163. 12:58Not has a really good one.
164. 12:59I wanted to use another one from you guys, but the notion is more reachable.
165. 13:04So here you see so if you have for example not every service has but the open API JSON what we doing with the protocol is translating this oh I already register it.
166. 13:22So, I'm going to show you guys.
167. 13:26Yeah.
168. 13:26Where is this one?
169. 13:29Sorry.
170. 13:34Don't see it here.
171. 13:39Okay.
172. 13:40It will form something like this.
173. 13:41This is like an example, but it's like the same.
174. 13:44So, you will get all the capabilities from the endpoints.
175. 13:49You see sometimes you see like a delete again.
176. 13:52So it would kind of be idea to approach like a middleware to this point.
177. 13:58Yes.
178. 14:01Oh, sorry guys.
179. 14:02Thank you for telling me.
180. 14:03And okay, here is better.
181. 14:09Oh, thank you.
182. 14:10Sorry guys.
183. 14:12Okay.
184. 14:13So the idea is to see here it's like you can translate the endpoints to capabilities.
185. 14:19So because we are still this is just because we still we don't have so much services or let's say none like implementation stuff but the idea is like in the meantime we have something
186. 14:31that translate this.
187. 14:32So what this mean?
188. 14:34It means like let me check.
189. 14:39It means like you will have something for example do you have there in the agent of protocol you see do you have a directory this directory we have for example
190. 14:54the Gmail one that we did the same.
191. 14:57So you see you have for example 20 list of capabilities.
192. 15:02So the idea is like you grab what the service can do, what is exposing maybe through open or maybe you do it yourself because uh for example maybe you want to automate a workflow.
193. 15:14So the idea is like everything get list here.
194. 15:17So the directory will be like a phone directory for the agent.
195. 15:21Are you following me?
196. 15:22Yeah.
197. 15:27The what?
198. 15:27Sorry.
199. 15:29Yes.
200. 15:29I mean it's a good point.
201. 15:31Did you hear?
202. 15:32Did you hear him?
203. 15:33Like fine grain out.
204. 15:34Okay.
205. 15:35So the idea is yes in a sense of having the most scope thing as possible but still fine out is still goes like minting a token to the user.
206. 15:47We want to meet a token for the agent.
207. 15:49So what we are switching is the principle.
208. 15:51We now want that the agents are a principal actor.
209. 15:55But it's a really good thing.
210. 15:56That's why we inspire on that.
211. 15:58I think it's one of the best things we can do.
212. 16:01Okay, great.
213. 16:02Um so going back to the capabilities the idea is what I just said like having this scope once you have like the readings or like maybe uh just the gets or the deletes or the post like everything is
214. 16:19quite discerned so you can decide whe you have any questions.
215. 16:28Yes.
216. 16:34Yeah.
217. 16:34Yeah.
218. 16:35That's that really sucks.
219. 16:36Uh the idea is you can do it manually.
220. 16:40You can actually declare everything so in the directory or in the future the agent know how to do.
221. 16:45It's like less common.
222. 16:47But yeah, it could happen.
223. 16:49But the idea is like in in the meantime we do this until services implement these sessions out.
224. 16:56But yeah, it's a good question.
225. 16:59[snorts]
226. 17:00Um okay so remember like our problems right so now we we are trying to get a solution for how like the agents the agents know what they do with the directory
227. 17:12then we have a way to authorize them through the capabilities and now if you remember we said it was a good idea to get them identity right so what we come
228. 17:26is giving agents like a private key like famous private key.
229. 17:32So each agent will have its own and that will be attached to the identity.
230. 17:38Um so in that sense because every agent has its own identity.
231. 17:44Now if you I see you want me to put this on light mode too.
232. 17:49Uh what did you say that before?
233. 17:52Okay, let's go back to this there.
234. 17:55That's better.
235. 17:56Great.
236. 17:59zoom in.
237. 18:00Okay, there.
238. 18:05So now that the agent will have like a access to its own private key, they can sign tokens, they can have like their own metered tokens and everything is encrypted with the private key.
239. 18:17So what this means like for now on instead of like a service let's say Gmail seeing like a user interacted with the service we can have a log that we actually say
240. 18:30okay this agent is doing something on my behalf that is connected to the user but it's actually for example agent from cursor or an agent from cloud or an agent from whatever.
241. 18:40So we are changing the paradigm.
242. 18:42We stop seeing as an agent is like hiding behind the user and now the agent is there as a principal acting.
243. 18:51So is is for me was really exciting.
244. 18:56So do you have any concerns about this private key private key public policy key things?
245. 19:02Yeah.
246. 19:09Exactly.
247. 19:10That's beautiful, right?
248. 19:12You can say, "Oh, this is not doing what I supposed to be doing, and I go and I revoke it."
249. 19:17So, it's cool.
250. 19:18And maybe you can think of like, okay, the the I revoke it, but the access token is still there.
251. 19:25So, what you can do is like deleting with the GDI.
252. 19:28So, bye.
253. 19:29Yes.
254. 19:38Mhm.
255. 19:39Always the token always the agents they always have a user they are reporting to you.
256. 19:43They are never detached.
257. 19:45So always the the agent is like with a user.
258. 19:51Yeah.
259. 20:01Yes.
260. 20:02Yes.
261. 20:02Exactly.
262. 20:03you have like all the information the agent also the host we are in introducing a new thing that is called host that is like the place when you are like creating this agent from so you actually can also delete the host you can say okay this host is not authorized
263. 20:18so you can see the host you can see the agent ID you can see the user you can see everything okay great more questions okay don't be shy like we are here for this
264. 20:32Yes.
265. 20:39Can you speak a little bit higher?
266. 20:55Yeah, that's a great question.
267. 20:57Um, actually it's more like an enterprise focus, right?
268. 21:00So because we did this in February, the things changed so much until now.
269. 21:06Like for example, Asians are not so ephemeral.
270. 21:08Maybe you have a forever long Asians, right?
271. 21:10Or maybe you have two Asians or maybe you have Asian design of a company and they have policies.
272. 21:16So we are working on that in a new draft and we are addressing that too.
273. 21:20Um because it's something quite um important to address.
274. 21:24So yes, I was about to say that in the end, but yeah, we're about to release a new draft with uh if you have this is an open source.
275. 21:30It's a project that we want to all of us own.
276. 21:33So if you have some ideas like you reach out like like that one for example, it's really good.
277. 21:38So we can improve it's good for ourselves and good for everyone.
278. 21:42And I don't want this to be just like for enterprise just to all of us that use AI that be safe.
279. 21:47So thank you.
280. 21:50Okay.
281. 21:51So that was that was just discussing about the V1 to V2.
282. 21:56Um great.
283. 21:58So three layers then one protocol.
284. 22:04Why we say this?
285. 22:05Because if you go and in the searching the website you will see we publish like a first version just like a draft or a ancient plugin inside of a draft.
286. 22:15The idea is you can use this and play around and see what you come up with.
287. 22:20So in the agent plug-in you can uh use it like on the server side so a server can verify an agent or issue grants or force whatever constraint that you have been doing with that
288. 22:34also on the SDK side uh we ship it with an MCP so you can connect it to your cloud to try it out you connect it to your whatever you're using course or whatever you're using with MCPS
289. 22:44and also the other layer is the directory it's just a place like a phone book like where the intent match what the action.
290. 22:55So intent will be matching the capabilities.
291. 22:59Um so the idea is that we can see it in action.
292. 23:07So what we will do if you want you can go to what I shared in the QR code.
293. 23:15Let's go to the page.
294. 23:19Okay, you will see a directory there.
295. 23:23Did you all find the directory button?
296. 23:25You will you will come into this and you will see a connect button.
297. 23:29So you have to login first for sure.
298. 23:31And you will see a connect and then you will see the MCP.
299. 23:36So you can connect that for example to the cloud.
300. 23:38I have it here.
301. 23:39I wanted to to show you guys a bit what we were talking about.
302. 23:44So let's say we say I'm going to ask for emails.
303. 23:50Okay, I just did a bunch of demo emails, but I'm going to show it to you guys so I don't show like personal information.
304. 23:56But for example, I have my NCP connected.
305. 23:59So you will see something like this uh connectors customize.
306. 24:05Here you see Asia D. You will see something like this.
307. 24:10Then you can approve, disapprove, always deny just the Asian D. um MCP and if it's that my idea is like you try it out just with MTP of Asian D. So the agent doesn't have access to other tools.
308. 24:23So it's like the principal one to see how it interact and how it gets the identity.
309. 24:28Our whole idea is like when you use this MCP the agent will have an identity we request you for capabilities and we'll try to execute it and you we will see the logs and everything
310. 24:37um uh running.
311. 24:39Okay.
312. 24:41So let's see for example um hey bring me my last email.
313. 24:49Oh sorry it's going to use the H&L connectors.
314. 24:56Can you see it's too small?
315. 24:59It's okay.
316. 25:01Okay.
317. 25:05Okay.
318. 25:05you see is using the because it has the NCP is using the directory to see from my intent of like bring me my last email which capabilities are out there
319. 25:16and he's asking for approval.
320. 25:18We are using here device uh device flow.
321. 25:21Do you know the device flow?
322. 25:22Are you familiar?
323. 25:24It's just like it's like connecting a device.
324. 25:26So we are acting like is a if the agent is another device.
325. 25:29So you can see uh the agent called is called email reader is that's for default.
326. 25:34You can change later if you want to.
327. 25:36It's requesting me to read at least my emails.
328. 25:40We're going to approve.
329. 25:44And I'm going to show you that guys later.
330. 25:51So the idea here um is that the agent will connect we get the identity it uh will have the last email.
331. 26:01Okay.
332. 26:04Okay.
333. 26:05We have a nice email.
334. 26:06So we know agents can read emails.
335. 26:09That's not the demo.
336. 26:11The demo is what what happens.
337. 26:13So remember we at the beginning we say if we gave identity we can have logs.
338. 26:17So let's see what the agent did.
339. 26:19So I know I have an active agent.
340. 26:22There's an email reader that has this agent ID.
341. 26:24He's in a local device.
342. 26:27It has these things and also I can see here this agent with this user ID.
343. 26:34Okay, I see the agent here just is the email reader and he got these results.
344. 26:39So I'm actually tracking down what the agent did, right?
345. 26:46So now let's say I want the agent to I don't know send an email by default what we are proposing is like all the hosts like for example clo
346. 26:58has reading capabilities by default so then it wouldn't mess up youration with your inbox right so let's see uh let's send an email who has like a really short email that I can put in there
347. 27:13hello really but no but actual one
348. 27:20[laughter]
349. 27:23who has like a good email.
350. 27:24If not I can put like my own but if you want to show okay so I gonna put mine you see I to myself saying I'm gonna use hello world.
351. 27:39I thought that your email was hello world was so cool.
352. 27:42[laughter]
353. 27:44Okay, so the idea here is if you as you just saw like the agent just have like the reading capabilities.
354. 27:56Ideally, it will never could send the email without me granting it.
355. 28:03So for example we can see it here right the only capabilities it has is two of these two.
356. 28:12So it will try to see it here.
357. 28:17We can also we also did like this extension for us to actually watch it better what the aation was doing.
358. 28:24Okay.
359. 28:25Okay.
360. 28:26Authorize patch with security verification.
361. 28:31Okay.
362. 28:32So you see this um the agent now is using CA like asyn like it's client back channel authorization instead of like just showing me like a redirect and see a device that I approve
363. 28:47I could approve from here.
364. 28:49So for the demo I just going to approve and now that the agent has approved.
365. 28:56I approve it.
366. 29:00It will be actually sending the email.
367. 29:04So you will see in the capabilities here that now you can send the capabilities of uh send emails and actually we can see here he just it just send the email right
368. 29:20so what do we want to try this we're going to see if you see no no okay here is the email okay so what we want to try now let's say this something went wrong with the agent Somebody was trying to access my Gmail trying to propector
369. 29:37whatever.
370. 29:38So the idea of this whole thing is to have the trustability we just saw is on the logs right and what if we want to revoke it let's try you let's try to revoke it and try to
371. 29:49for example read my emails.
372. 29:51So we're going to revoke it first so it cannot see a thing.
373. 30:00Oh I lost the page here.
374. 30:03[snorts]
375. 30:03So in agents we can revoke the agent.
376. 30:07So bye-bye.
377. 30:09And now we can say read that last email you just sent.
378. 30:18Okay.
379. 30:26You see the connection got revoked.
380. 30:30So what happened now?
381. 30:33Why it can't read it anyway?
382. 30:36Because what we did is like he created another agent an identity because I revoked the one before but this is just happening because the get the list is a default capability in our host because for us it's safe.
383. 30:51What if now I wanted to say um send an email saying hello to myself.
384. 31:03It will do the same.
385. 31:04It will ask for it.
386. 31:05So actually it's working right.
387. 31:07I can actually revoke it.
388. 31:08I can see the logs.
389. 31:09I can say what the agent can do or not.
390. 31:12Don't you think it's a it's kind of nice to have this kind of workflow?
391. 31:16It's more controlled.
392. 31:18Yeah.
393. 31:19You have any questions about this?
394. 31:20How something that pops in your mind?
395. 31:22Yeah.
396. 31:25Yes.
397. 31:26Yes.
398. 31:27Yeah.
399. 31:28Every constraint that you want to be careful about, you can put in the capabilities.
400. 31:35So, it's good.
401. 31:43Yeah.
402. 31:45Yes.
403. 31:49Yes, exactly.
404. 31:50You can say maximum maximum time of executing.
405. 31:53You can say the time.
406. 31:55You can say everything.
407. 31:56Yeah, it is.
408. 31:58Thanks there.
409. 32:11Mhm.
410. 32:19Well, you can have both.
411. 32:21Our idea is like to be safe but not to be annoying.
412. 32:24Imagine every time you had to read something.
413. 32:26Oh, I had to approve every time.
414. 32:27You're going to hate us, you know.
415. 32:29So the idea is like a host for example or inclus sorry I spoke in Spanish include um or like maybe a host like someone specific you want like for example let's say co can do everything and then like are recently joined just have like this specific access
416. 32:47so you can do it like by user policies as I was talking about or you can do it like host policies.
417. 32:52So you have to be like in the middle I think of like good UX uh the good user experience but also secure enough
418. 33:00but ultimately
419. 33:04yeah is the ser yeah yeah yeah because uh agent what we're doing like as I said you can try it out uh with the server part is the one going to verify the agent like back with the users going to verify the policies we're going to verify has access or not so yeah all the time verifying so
420. 33:24Great.
421. 33:25Um, so the whole idea was to show you this.
422. 33:30Uh, I was trying to ask the to send.
423. 33:33Um, I need to confirm.
424. 33:34So you see, so I here I have the request.
425. 33:37I didn't accept it.
426. 33:39So let's say I deny it.
427. 33:40It will be like the same.
428. 33:42I just did.
429. 33:45So ideally he wouldn't.
430. 33:50You see that's red.
431. 33:51Okay.
432. 33:51So it's working good.
433. 33:53So he wants me to allow it and I'm going to I'm going to So can you tell me agent status?
434. 34:00So you can have all the time like track down what agent is doing if it's revoked or not.
435. 34:07Yeah.
436. 34:24Oh, are you saying if you have for example a skill that is saying to use the MCP server differently
437. 34:32directly?
438. 34:34Yeah.
439. 34:34Well, actually what we are doing is we are infor in this just an example for you to see how the protocol is doing just like a demo but you can enforce for example that the agent the only thing that can use
440. 34:46um is the MCP tools in this case that's what we did with the plug-in but yeah you have like for our next protocol it's going to be like more
441. 34:57uh more cap down so the only way the agent can go out and ask for services is through the agent out protocol so it's a good really good question.
442. 35:06Yeah.
443. 35:09The what?
444. 35:10Sorry.
445. 35:12The proxy
446. 35:16here.
447. 35:24Yes.
448. 35:24Uh you have this in the directory.
449. 35:27Now you can use it like when you log into for example you are connecting a lot of services in the directory.
450. 35:32So the agent when it connects to the MCP can know all the capabilities that it can do.
451. 35:36So inside of the MC the proxy you can see for example the Gmail you can see all the capabilities that agent can do.
452. 35:42You can see like the agents doing stuff the logs and everything.
453. 35:46So every log is like cut down by provider.
454. 35:50Okay.
455. 35:53Okay.
456. 35:53Um and about that um our actually proxy is just a directory.
457. 35:58It's not a real proxy for us like proxy is a really bad idea.
458. 36:01It's not scalable because proxy uses data.
459. 36:04So in this case it's just like a directory matching intent with capabilities.
460. 36:11Okay.
461. 36:11Um so so back again if you want to try it you can like live you can use like the plugin the SDK.
462. 36:27It's really nice because we started like from the problems and now actually the solutions right like the discovery authorization and identity.
463. 36:36Um this is kind of a what I will show you guys like MCP how to connect it.
464. 36:42Um the SDK side with the MCP is the one that is creating the the the keys and assign it to the to the agents and the server side is like verifying
465. 36:53authorizing grant um giving the grants that are authorized by the user and and so on.
466. 37:00So for me the most interesting thing of all is we need to stop giving credentials our credentials to our agents we need to give them authority this should we stop saying like pretend to be me
467. 37:19instead of saying like act for me within these limits.
468. 37:26So we are almost in the end.
469. 37:29If you have any questions, something that pop into your mind or it could be also an addition of what I'm saying.
470. 37:36I don't know if you ever thought of this.
471. 37:37Do you think it's a good idea that we implement something like this?
472. 37:40I'm not saying this one, but maybe.
473. 37:42Yeah.
474. 37:44Okay.
475. 37:45The idea is that this is open for everyone.
476. 37:47So I would love you guys to join us on Discord or you can send me a message, send me an email, be in the channel like be active because if we improve this for all it's going to be good for all of us.
477. 38:00Yes.
478. 38:06Yes.
479. 38:12Yes.
480. 38:12Yes, please.
481. 38:13And please do.
482. 38:14Yes, please do.
483. 38:17Sure.
484. 38:25No, it's totally diagnostic.
485. 38:27Yes,
486. 38:32exactly.
487. 38:32That's the idea.
488. 38:42Yes.
489. 38:47Yeah, that's on our B2 and that's our prime focus now.
490. 38:50So, yeah, that's a great idea.
491. 38:52Great.
492. 38:54Any anyone else?
493. 38:55Yes.
494. 39:00No, it's different.
495. 39:02Yeah.
496. 39:03No, we also I mean we have like as an inspiration like everything that's out there, but it's not the same.
497. 39:09It's different.
498. 39:10Yeah.
499. 39:11Because they are not still treating ancient as a principle.
500. 39:14We are treating nations as a principle.
501. 39:16They have their own identity.
502. 39:17So from that it's different in that way.
503. 39:26I don't know.
504. 39:26We don't I don't have that information.
505. 39:30Yes.
506. 39:35Sorry.
507. 39:35What?
508. 39:40Oh yeah.
509. 39:42You see I it's connected because of the discovery but it's still the same like for me the most important part is like the Asian B principles have their own identity AI gateway
510. 39:55doesn't have doesn't do that yet I hope it does but the whole idea is to have like traceability have the whole life cycle and to be able to track down and
511. 40:06hunt if you want an agent so that you only have if you have identity on on each agent.
512. 40:12and AI gateway is not providing that.
513. 40:16Sure.
514. 40:17Is there any more questions?
515. 40:21Okay.
516. 40:22Well, you can reach me in the that's my email if you want to email you want to contribute together would be nice.
517. 40:29Uh that's my link in and the discord channel from the whole Asian protocol.
518. 40:34Okay.
519. 40:34Thank you so much.
520. 40:35It's been a nice one.
521. 40:37[applause]
522. 40:51[music]