WebDev

@30150821474450 for ivfindia: assigned 5 bugs with ETA as today. 3 bugs with eta as tomorrow for reference: https://docs.google.com/spreadsheets/d/19oi1dL1PTsPrzYj7tEVM9S5xjsvQTewNEYiPFjimqbI/edit?gid=0#gid=0 you can also check your queue for support page

@159738440466644 @54125194641609 GOOD work on sitemap task. much appreciated 🙏🏻🙏🏻

The crawler/scraper can be configured by a god on https://beyond-chats.netlify.app/beyondchats/god/crawl-setup

Flip this toggle to “on” to run the process weekly for the client.

The much awaited sitemap scraping is live now. It is a god only feature for now and we plan to publicly release it in the next sprint (by 17th august). IMPORTANT NOTE: WE MUST ONLY USE IT FOR CLIENTS WHOSE sitemaps DO GET UPDATED. What is can do: - Automatically fetch sitemap. If it cant then we can manually enter the sitemap link. - ⁠Find all new/deleted/updated pages of the client using their sitemap then train the chatbot on the selected pages. We also plan to run this automatically on sundays (at 2AM IST) every week from next week. It will be run for the clients we select. Some known …

please check the other properties as well (like expanded) incase that is causing an issue

updated

sure

E2E_TEST_ORG_ID=beyondchats

Can you please modify this one variable E2E_TEST_ORG_ID to “beyondchats”, id isn’t working during a dryrun that’s why. Thanks.

Sure also hit a blocker when creating staging mrs for that task, laravel MR is fine but for dashboard in staging these changes are missing refactor/standardize-integration-pages

@30150821474450 hey i will be assigning some bugs. Can you work on those after staging MRs for integration page fix?

they are added

@111561876209776 can u please confirm this once.

@159738440466644 would be great if you can share a demo video today of how playwright testing will look like

@analyst start staging server

@analyst start dev server

Progress today & next steps

Can you please add these variables.

Please add the variables I sent you, to beyondchats-embed repo ci/cd vars. Thanks! @111561876209776

shared my username in your dm. please do check.

@159738440466644 are you on discord? I will add you to our discord server where the errors/crashes get reported to

yes

@111561876209776 should i start working on this, or else anything other needs to be done. please do let me know

Yes! Please do let me know if there's any task that needs to be done beforehand.

Thankssss ::)

Here is the full codebase for Questionnaire popup

<media>

@117111057871026

<media>

Nope It's a toggle window which pops out . I am pretty sure it's codebase is somewhere else

@117111057871026 Is it not available in DrMalpani codebase?

@159738440466644 hi did u start work?

i will have to check

@111561876209776 Do you guys have this branching json file for this questionnaire somewhere, like what you actually use here. Can you share it with me.

@analyst start staging server

@analyst start dev server

EOD Update Fixed the connected status is broken on integration pages issue you raised What was happening Integration pages showed Connected as soon as an account was linked even if it had stopped working. So WhatsApp bulk messaging could be failing while the page still looked healthy. The uptime monitor knew, but that page is internal-only. What's fixed now Every integration page now shows the real state, Connected, Needs attention, or Not working and tells the user what to do next with a button to fix it. No more raw error codes shown to customers.

<media>

<media>

<media>

<media>

<media>

@30150821474450 EOD updates?

<media>

<media>

updated task sheet for this task

not yet. I will let you know

Have you had a chance to review them? Please let me know if any changes are required, and also about the next phase. Till then, I'll continue working on this task.

yea for past 2 days laravel was really fast

*Staging MR:* https://gitlab.com/dev_management_beyondchats/chatgpt3_backend/-/merge_requests/1731

added my pc as a 2nd gitlab runner so CI jobs should be way faster now

In the mean time, please do have a look on this: https://gitlab.com/dev_management_beyondchats/chatgpt3_backend/-/merge_requests/1697

sure, please give me few minutes.

@159738440466644 i want to take a look at the unit tests u've created till now. Please either create a MR to staging for the work done till now or share a document with code references

Good

That was for my understanding so I can check on my phone never interested in such long pdf

Easy to read, easy to understand, easy to give feedback on

@202611475136752 Such simple, yet clear plans are MUCH better 52 page documents.

Please call me when you start work

sharing doc link here, so that any modifications, requirements or plan improvements can be added here: https://shorturl.at/ZeCOh

<media>

why is it a waste?

sharing a doc in few minutes.

5 din m saare compute credits khatam. Strange

What’s the imlpementation plan / what are the various phases? What are prioritising first?

Worked on phase 0 unit test implementation for Laravel backend. There are still many changes//enhancements needs to be done. scope is vast. will now share updates about this regularly.

@111561876209776 Let's discuss this once. I feel we are wasting time on this.

<media>

1. Added a new section for non-API endpoints, including empty and unmatched routes. 2. Started working on the Laravel backend unit test plan for Phase 0 (using PHPUnit over Pest, which has some trade-offs).

If they forget to do this, you need to nudge them. And if it still happens again and again, we should take more serious actions.

So we know what they’re upto, how are things progressing

@111561876209776 Yar khud se thoda dhyan do. Each team member you supervise needs to share verifiable, updates in the group every few hours.

😕

@159738440466644 Not a single update yesterday?

@analyst start staging server

@analyst start dev server

https://drive.google.com/file/d/16mWTjzPEZo0kbfzkZ_fAiG2FJFrZSbsb/view?usp=sharing Added Shadow DOM for embed in dev, to differentiate before and after, built a test lab https://dev.beyondchats.com/embed-test/?env=dev&profile=clean&org=ivfindia.com , where nuclear option in the dropdown injects css from the host site to the embed also few other breakage options can be tested by selecting in the dropdown , u can switch between dev/staging/prod to see how the embed looks with the shadowDOM fix (currently only in dev) and without it when the breakage occurs

@analyst start staging server

Gated Resolution for support page

https://drive.google.com/file/d/1GjhzQWNfy-XE8gTJbUMSNK3Z1RSZqHym/view?usp=sharing

thanks, video demo is helpful

Progress today: - staging MR for newly added chat filters. - tier override dialog with radio cards for Core/Major/Minor - fixed minor tier filter showing zero results (null tier fallback) - routes now show with leading /, unmatched requests moved to separate info section

<media>

<media>

<media>

Great ux additions. My bad I missed all of them!

Also, there was a bug where new convo button on the header stopped working when on chat history page. Fixed this.

- ⁠added “raise a ticket” to the starter page

@159738440466644 made some small changes to dashboard chatbot: - added minimization so it doesnt cover dashboard content

<media>

every metric needs to be hyperlinked. If I click “99” -> it takes me to the support ticket section with relevant filters auto applied

ok will add it as a dialog box

are we gonna show this form on the reply box itself?? it needs to open as a dialog after someone clicks “mark as resolved”

<media>

no the bug fix report

<media>

this task has report??

please get things verified first

how does the report look like?

Creating staging branches now

It is done

@30150821474450 need this completed today, take it up right after the bug report

@all Please run claudelens:sync-history feature after updating your plugin. It is needed so that updated information and fields are being sent by your claude code to the server.

Dont send to pankaj@beyond Send to contactbeyondchats instead

We need to update the claude.md?

claude needs to know where to insert what BIG TIME otherwise our tech debt will keep increasing

Oh ok will ad it there then

https://beyond-chats.netlify.app/beyondchats/email-management

it was made live a few weeks back

we have an /email-management page now

For god users

In the support section itself?

Ok where to add it in the dashboard 🧐

Irs for a daily bugs resolved report cron

need the emails handled via the dashboard and not hardcoded

correct

crt kya hai

devanshusinghai2005@gmail.com siddhantjain209@gmail.com pankaj@beyondchats.in @111561876209776 could u confirm if these are the crt email ids ?

Should I raise staging MR for this, or else any modifications are required. Please do let me know. Thanks.

Thank you so much for filling the gaps. My bad, i didn't check them before finalizing. It had many edge cases actually, probably that's the reason I got carried away. Will definitely rectify this from next time.

Filling it.

already fixed both these issues

@159738440466644 the “ask user for email after N messages” was vastly incomplete. - no hiding “continue as guest” button if the value of ask_email_after_messages was 0 - ⁠clicking on ground truths if ask_email_after_messages threshold was already crossed was crashing the embed project.

Yes, I have added the link in task updates too

Is it added in product sheet also?

also, @159738440466644 every AI/ML related pipeline needs to be put on beyondchats-node or background-flask project and NOT LARAVEL. I have the updated the architecture for dashboard chatbot but please follow existing patterns moving on

task updates not filled for this task

@30150821474450 you recently created a MR for: “Chatbot keeps starting each msg with users first name" right? Please update the task updates column for it

@102095734710467 @267155589353519 emon has created a guide for how to scrape a website using openclaw: https://app.notion.com/p/Open-Claw-BC-scraper-3aea4e78beb9801bbabefbf2e7193186?source=copy_link I havent yet reviewed the guide

@all, please follow this on your PCs. Thanks

soon you will be able to ask these questions to the dashboard chatbot once its been made live

row 30 of guidelines subsheet

hum bhul gye apne

dev server ke creds kaise set kre

I think we should just add a timer to this command: “@analyst start staging server —timer 4h”

Dev server and staging server have scheduled downtime from 9 pm to 9 am

If someone is using the dev server they will definitely call some endpoint, so if all endpoints were called >1hr ago then the server will stop

I’m using it right now

I feel we should make this way that if someone uses dev server and dev server is inactive for 1 hour it just stops

Should we keep dev and staging running 24x7 ya this is fine?

@111561876209776 is this a nuisance?

@analyst start staging server

@analyst start dev server

@analyst start dev server

And after that please run /claudelens:sync-history command , you will get the option to backfill all the projects session that are shown on dashboard. Kindly follow the steps accordingly and do it as well

Please updates your plugin ASAP on claude code

@all This is the new dashboard for claudelens at https://claudelens-dashboard.drmalpani.com. Do check it out.

When do we plan to make it live?

@111561876209776 Dashboard chatbot status?

Got it will check it out

adding this to task details of existing bug

there needs to be a single source of truth which is the uptime monitoring please

<media>

@30150821474450 another case of: “connected status is broken on integration pages”. The uptime monitor is correctly showing that the meta whatsapp integration is down but we are still showing "connected" on the actual integration page

1. they will need to redo the integration setup. We have lost access to their waba, they might have removed it from their whatsapp manager 2. ⁠client-owned WABAs onboarded through embedded signup never appear in our Business Suite asset list. For example: apnipathshala doesnt appear either

@111561876209776 update on this?

Ok 👍🏻

<media>

@30150821474450 please move the copy button below the actual message

<media>

@102095734710467 are we good to go, with these filters. please once do try them on dev site as well: https://dev.beyondchats.com/dashboard/ivfindia.com/68795

okay

And delete faaltu numbers from Met

@111561876209776 Please share where I can see this number on meta dashboard

<media>

Call me if you need signin link / OTP

Hi everyone, including @54125194641609 & @117111057871026 beyondchatss@gmail.com is again on 20x usage tier of Claude Code. Feel free to use that moving forward. But please be mindful of your per-session usage. Doesn't mean you need to reduce your productivity by using crappy models in low effort, but just means you need to learn which model and effort is "good enough" for your use-cases.

@analyst start dev server

Latency depends on tea breaks

AI (Actually Indian)

Semiconductors are expensive. Let’s build LHMs instead. Large Human Models. People send msg to a chatgpt like interface. Large centres with human agents reply.

BeyondChips

Better to start with semiconductor industry

GPUs are old, we will make our own neural engine, step one, gather sand

Then our own GPU’s

i feel we can save our aws costs if we just make our own aws

<media>

Let’s make our own LLM 😅

Will do some r&d around

Agreed, the concepts are solved point is more about implementing them effectively in our workflow.

These are solved problems.

It’s because we’re still doing heavy development work, so it’s naturally more exhausting. The biggest thing we need to figure out right now is the best practices and workflows that everyone can follow and use efficiently.The main pain point is maintaining product memory end-to-end. Even if we need to update or modify a feature that was built months ago, the model should be able to understand the context without consuming a huge amount of tokens every time. That’s probably the biggest challenge for us right now.Once we solve that, the actual development becomes much easierwe can use highend mod…

@202611475136752 bro leave some tokens for me , every time i try to use it is at its limits 😞

There is a reason why Claude has provided so many options

And learn to use the right models, at the right efforts, with the right tools for different tasks

Koi baat nhi. Jo hua so ho. But team, please be mindful of your usage patterns and spending yar.

@202611475136752 m

Nobody else has creds for support except you and Yash

Yash, I think

One

support@beyondchats.in

@102095734710467

Who do i share mine with

I share it with Devanshu. And we rarely hit our limit because we are mindful of which model we use, at which effort, for what purpose, etc.

Sbse zyada mere account se use hota hai Claude

dev team

19$one

devteam

Which one is that? 🥲