FIXED 1 BUG. CREATED 3 MORE.
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
A wide-ranging working stream focused on reviewing a Shopify app before launch, combining chat and merchant-UX refinement, fault diagnosis, and conversion of review findings into a debugging and testing queue inside a custom command center. Practical work covered crashes caused by products without images, disappearing text or attachments, slow replies, alignment problems, contextual actions, and confirmation states. A recurring conclusion was that a local fix is not proof of system readiness: human review and realistic scenario testing remain necessary. Luqman also explored agent-driven debugging through backend tests and log inspection, context management through documentation, state files, and fresh sessions, plus confirmation gates and Git checkpoints to constrain autonomous changes. Audience questions led to substantial discussion of Claude, Codex, OpenCode, OpenRouter, local models, and agentic coding tools. The practical view was that model choice depends on the task, quality, cost, and familiarity with the project, while benchmark tables must be interpreted through their methods rather than a single ranking. Many multi-model and cheaper-routing ideas remained proposed experiments rather than completed findings. The stream also covered validating product ideas with landing pages, taking products to market instead of repeatedly building, and moving from a useful first release toward scaling based on real usage. Visual brainstorming and interface design, SEO and GEO, precise product descriptions, and an Arab knowledge-sharing community were additional useful threads. Open questions remained around some latency and timeout causes, classifier-context cost trade-offs, automation reliability, and several fixes and tool trials that were not conclusively validated.
155 moments from the stream
- 7:25:27
A Limited Context Window
The speaker observed that the application’s context window was short and that the assistant forgot earlier events after only a few messages, weakening conversational continuity.
- 7:26:10
Model Transparency and Merchant Perception
The speaker favored disclosing the technology powering the product on its marketing page. He expressed the opinion that North American Shopify merchants might prefer Claude over DeepSeek because of perceived quality and origin, presenting this as a marketing judgment rather than a documented test result.
- 7:30:16
Balancing Conversation Memory and Cost
He explains that a reply can be immediate when the context is retained; otherwise, the system searches again from the beginning. Expanding the prior-message window is possible but substantially increases cost.
- 7:30:55
Starting Project Evaluation with Profitability
An initial suggestion for evaluating a new project: begin with expected profit and first ask whether the idea is economically viable. The detailed evaluation criterion remains incomplete in the captions.
- 7:33:19
Closing Chat and Task-Board Work
After reaching a satisfactory state with the chat work, he parks that area for now and marks the task-board and assistant prompt-card tasks as complete.
- 7:34:55
Reviewing a Privacy-Focused, Pressure-Free Prayer App
A positive review of a prayer-app concept combining accurate times, qibla direction, privacy, and prayer statistics, while distinguishing its experience from apps built around urgency, gamification, or psychological pressure. The discussion also asks how a free app would be monetized, separating monetization from influencer marketing.
- 7:38:48
An iOS Prayer-Time App with Siri and HomeKit Integration
A discussion of an iOS tracking app that visualizes prayer times, lets users ask Siri for the next prayer, and can trigger a home scene based on prayer time. The conversation also raises, without resolving, whether its sustainability model would use subscriptions or open source.
- 7:44:24
Reviewing UI Issues and Tracking Task Progress
A UI review identifies misleading timing for quick-action buttons, too many stacked page sections, and a need to restore a branded dark hero card. The calendar and task list are then checked to distinguish completed work from remaining tasks.
- 7:50:07
Turning Floating Assistant Access into a Standalone Feature
The assistant was found to be reachable only through a menu link. Floating access across sections was therefore treated as a standalone new feature and sent through a full preparation flow: research and exploration, creation of a handoff file, and then implementation by a builder agent.
- 7:55:07
Reviewing Progress and Checking Agent-Generated Tasks
The calendar is used to review accumulated debugging progress, with a manual debugging pass planned around a checklist. Some task details added by the agent are also flagged as inaccurate and in need of human review.
- 7:56:32
A Customizable In-Project Command Center Template
A plan is described for a Command Center delivered as a Markdown file or prompt template that users could place in a project, then adapt and polish to their own needs. It is presented as something to share and explain later, not as a completed result in these pages.
- 7:58:56
The Trade-off Between Stream Latency and Quality
He explained that live-stream latency could be reduced, but doing so might lower stream quality, creating a practical trade-off between faster interaction and visual quality.
- 8:00:36
Reviewing Clickability and UI Element Positioning
He identified two interface issues: an element was not clickable and sat too low near the bottom. He later described it as clickable, then tested clicking it and said it took him to the assistant chat.
- 8:01:09
Combining a Main Work Stream with a Silent Stream
He discussed splitting daily streaming and leaned toward a longer morning work session, with a possible quiet evening stream using text chat instead of spoken commentary. He presented it as an idea under consideration, not a confirmed plan.
- 8:03:25
Shipping the Product Before Exploring Automation Tools
Responding to a question about agent and automation tools, he said his knowledge of some of them was limited and that learning them would take time. For now, he preferred avoiding distraction and focusing on shipping the product, while potentially revisiting marketing automation later; he also expected automation broadly to advance significantly.
Got a question about something in this stream?
Ask in the group