NANO 🍌 BUILT THIS
Want to back the stream? The «داعم» membership on YouTube is EGP 60 a month: badges, emoji, and I answer your questions first on stream. Become a supporter
The gist
The stream follows development of TalkStore, a proposed AI assistant for Shopify merchants, from shaping a pre-launch waitlist page to testing access to product, order, and theme data and building GraphQL wrappers with personal-data safeguards. Luqman discusses specialized research, build, review, security, and compliance agents; usage limits that constrain parallel work; store-context caching; and conversation persistence. Practical work covers deploying an Astro frontend through Cloudflare, connecting the domain, Supabase, and the waitlist form, and improving social metadata and lightweight analytics. Deployment and DNS pass through repeated errors and pending states before later progress is reported. A code audit surfaces significant authorization, Prisma connection, expired API-version, and GDPR-webhook issues; some fixes are reported, while follow-up remains. Audience discussions also cover video performance, model routing, OpenClaw, code security, freelancing, Shopify basics, and distribution. Final Shopify approval and compliance remain uncertain, as does the measured benefit of some caching work, while the broader discussion argues for beginning marketing and waitlist building before launch without treating the overall build or review as fully finished.
119 moments from the stream
- 2:41:08
Suggested Platforms for Starting Freelance Work
In a brief response to a question about starting a freelance business, he points to Upwork and Fiverr as platforms to consider.
- 2:41:37
Testing Analytics Questions and Completing the Orders Task
After not finding a direct analytics section, he tests a question about the best-selling product and revenue over a recent period, then marks the orders task complete. The store’s actual result is omitted.
- 2:48:25
Diagnosing a Failed CLI Redeployment
He suspects a Git-triggered build using an older dashboard, then explains that redeploying through the CLI still returns an error despite supplying its output. The attempts continue without a claimed resolution.
- 2:51:06
PII Guards Before GraphQL Execution
A task description proposes two guard functions: blocking filter tokens associated with personal data and recursively scanning outputs for sensitive field names, while explicitly listing allowed filter keys and warning against PII use.
- 2:52:42
Diagnosing a Deployment Loop Across CLI and Cloudflare Dashboard
A discussion of project setup and deployment repeatedly failing, comparing CLI and Cloudflare Dashboard deployment while questioning why the workflow changes deployment methods instead of resolving the original error.
- 2:55:19
Practically Reviewing Email-Filter Protection for Orders
A practical review asks for orders associated with a customer email, and the resulting response refuses filtering orders by customer email. The discussion also raises—without resolving—whether a general request for paid orders from the current month relates to the guard system.
- 2:56:49
Meaning of the Colloquial Phrase “I’m Your Uncle”
Explains that the phrase is used figuratively to claim greater experience or understanding, rather than literal kinship.
- 2:57:04
Pausing a Security Agent Over a Token-Usage Issue
The speaker says he is not currently using the security agent because of a token-usage issue he attributes to Anthropic; this reports his current situation rather than a documented test of the issue.
- 2:57:54
Preparing Two Development Tasks to Run in Parallel
The speaker prepares two development tasks to run in parallel for the first time and expresses frustration with deployment problems encountered with Cloudflare and Railway; these units do not show the tasks’ outcome.
- 3:01:22
Persisting Conversation Metadata and Sequencing Development
The plan calls for persisting conversation messages with domain, verb, model, token counts, and task-consumer flags, then moving from the current milestone to a section-generation engine and a theme duplication, installation, and preview flow.
- 3:03:16
Caching Shopify Theme and Product Context
A proposed per-shop, layered cache fetches theme context from Shopify on a cold request and pre-injects it into warm system prompts. It also compresses a sample of products and metafields into a snapshot shared by domain handlers. The goal is to reduce tool calls, latency, and token use while deduplicating concurrent fetches; no final measured result is shown.
- 3:08:44
Risks of Running Development Tasks in Parallel
The speaker monitors two development tasks running in parallel and notes that errors could indicate a clash between them. Errors appear and changes involving theme context, caching, a database, and the Shopify API are reviewed, but the captions do not establish a successful final integration.
- 3:13:48
Caching Store Context in the Agent
He explains that caching details such as the theme name, fonts, and colors lets the agent answer later without calling tools every time, potentially improving speed, reducing token use, and producing more specific responses.
- 3:16:50
Deploying a Landing Page and Connecting a Custom Domain
He follows the landing-page deployment through Cloudflare, then starts configuring the custom domain and locating the DNS settings in GoDaddy. The captions indicate that deployment completed while DNS setup was still in progress.
- 3:19:48
Trying to Verify the Cache’s Speed Impact
He tests a question that the cache should answer quickly, but the response times appear similar, so he cannot determine from the visible test whether the task was implemented correctly.
- 3:23:05
Inspecting an Unexpected Local Prisma Studio Page
He is surprised to see Prisma Studio at a local address and questions how it relates to Supabase before considering two options: a quick fix or debugging first. The captions do not establish a cause or final resolution.
- 3:23:56
Connecting the TalkStore Domain Through GoDaddy and Cloudflare
The discussion traces custom-domain setup through CNAME and root-domain configuration, confirms that DNS was still managed by GoDaddy, encounters a record conflict, and then begins replacing the nameservers with Cloudflare’s. The segment ends while propagation is still pending, not with confirmed completion.
- 3:25:26
Practical Review After Completing Development Tasks
The speaker explains that a task may be completed before a deep review, with a later security sweep, debugging, reviewer checks, and extensive testing. He prefers feedback grounded in practical experience of the result, while noting that webhook wiring was deferred to a future task.
- 3:27:15
Testing an Injected Product Snapshot in Agent Chat
The segment discusses backend logic that injects a product snapshot into the chat without a visible UI change, intended to let the model know product information without rereading it each time. Testing was proposed through messages in the embedded app, but the described attempt showed no noticeable speed improvement between the two questions.
- 3:42:07
From Localhost to a Domain as the First Shipping Step
The speaker describes moving the site from local development to the TalkStore domain as the first shipping step, while stressing that the work was not yet complete and still needed polishing before a pre-launch campaign. In the historical timeline stated, this was discussed on April 1 against work planned through April 12.
Got a question about something in this stream?
Ask in the group