RIP. NEW MISSION. CC V2.0
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 centers on redesigning a largely autonomous “Command Center” that routes coding tasks by size and dependency, delegates research, exploration, and implementation to agents, and introduces task- and milestone-level audits before human review. Luqman demonstrates unfinished practical work across Codex, Claude, and terminal workflows, weighing output quality against token consumption and usage limits while encountering real friction involving Supabase, Docker, MCP, and stalled tasks; the workflow is therefore not presented as a completed success. The discussion broadens to unit testing, lightweight feature specifications, competitor research before building, and the need to market products rather than stopping after implementation. He also outlines a Shopify tool concept for measuring and improving product visibility in AI answers through GEO, which he described at the time of the 2026 recording as an emerging field and forecast to grow, particularly toward late 2026, while leaving market viability, implementation complexity, and the GEO–AEO distinction unresolved. Drawing on experience that began with dropshipping in 2017, Luqman compares dropshipping with DTC and white labeling, e-commerce with app economics, and B2B with experimental B2C directions; he also offers opinions on SaaS margins and recurring revenue, reasons for choosing the Shopify ecosystem, and app marketing through defined audiences, acquisition channels, and a mix of organic and paid approaches, without presenting these views as guarantees or settled outcomes. Secondary discussions cover automated short clips and the risk of lost context, recorded streams as a work archive and community-learning channel, programmatic marketing videos with Remotion, thumbnail experiments, and copyright constraints. Audience questions add useful discussions about learning software development through practical AI-assisted projects, verifying the security and provenance of agent skills, estimating time, managing dependencies, model pricing and compute efficiency, and scheduling work around session limits. The qualified conclusion is that automation may reduce manual coordination but does not remove the need for oversight, debugging, and validation; Luqman argues that strong planning and context matter more than minimizing token use alone.
155 moments from the stream
- 6:34:52
Discussing the Motives Behind Anthropic's Limits
Responding to a characterization of Anthropic as greedy, the speaker acknowledges a problem but doubts that the intent is to inconvenience or exploit users, calling for a more reasoned view of the company's motives. This remains his personal opinion.
- 6:37:32
Does the Workflow Need a PRD or a Thin Feature-Spec Layer?
The discussion compares a product requirements document with the existing strategy and tracking files. It suggests that the missing piece may not be a traditional PRD, but a concise per-feature specification linked from tasks to clarify user intent and approval decisions before implementation.
- 6:44:25
Designing the Feature-Spec Layer Before Implementation
When implementation of the feature-spec layer is proposed, the response notes that real design choices must first be resolved, including storage location, template shape, tool surface, and approval flow. A brief design is therefore proposed before writing code.
- 6:48:09
The Role of a Task-Expansion Agent in an Agent Architecture
The discussion considers whether a task-expansion agent fits the current architecture. The qualified conclusion is that there is a plausible slot and existing hooks for it, but its impact could range from a natural extension to a load-bearing change depending on what “expansion” means, with token efficiency as a central concern.
- 6:53:38
Routing Tasks by Token Consumption
A routing idea assigns coding tasks between tools according to their expected token consumption: heavier tasks go to one tool and lighter tasks to another. It remains a proposed strategy; no validating test is shown here.
- 6:54:36
Validation Blocked by Missing Local Docker
A task remained under review because Docker-based validation could not run on a machine without a local Docker binary, so the task could not yet be treated as complete.
- 6:56:56
Separating Intelligent Planning from Agent Task Execution
The speaker argues that much of a task’s quality comes from the instruction structure and system prompt given to the agent. He therefore prefers strong reasoning to produce a clear “map” or prompt, after which another model may handle straightforward execution, accepting additional token use to produce prompts that are clear and comfortable to work with.
- 6:58:04
Following Docker Setup and an Outstanding Review Task
The page follows a Docker-related setup step until a success message appears, then surfaces an outstanding task still in review and asks what remains to be done. The captions do not provide enough detail to identify the project or the exact setup steps.
- 6:59:11
Routing Task Groups to Alternative Models Through OpenCode
The speaker describes a command that gathers tasks of a specified level and prepares them together, while noting that this can consume many tokens. He considers routing that work to connected models through OpenCode, but remains hesitant; the captions do not fully clarify why.
- 7:01:06
Choosing Coding Tools Around Available Usage Limits
In response to a question about models for OpenCode, the speaker discusses choosing coding tools such as Codex and Claude and making practical use of the usage allowance available in each session. The point is presented as a personal workflow preference, not a benchmarked model comparison.
- 7:16:43
Working in the Morning for Better Productivity
The speaker says starting work in the morning helps him accomplish more and that long nighttime streams are harder for him. This is presented as personal experience, not a general rule.
- 7:18:21
Caffeine Timing and Its Possible Effect on Sleep
The speaker relays advice from Gemini to avoid coffee after noon because caffeine may remain in the body for several hours, while acknowledging that effects differ between people and that he is unsure about some details.
- 7:19:15
Recorded Streams as a Work-Progress Archive
The speaker explains that recording his work lets him revisit project progress, his reasoning, and the tools he used. He also sees the interaction as a useful two-way learning exchange with the audience.
- 7:24:57
Confusion About the Codex Usage-Limit Indicator
The speaker observed that the Codex usage indicator dropped after a task and later appeared higher, contrasting it with Claude's reverse-style display. He presented this as confusing behavior without establishing a cause.
- 7:29:05
Codex MCP Integration Failure
A message indicated an MCP failure with Codex, after which the speaker described Codex as broken and tried to continue with another setting or alternative. The fragmented captions do not establish the cause or a confirmed fix.
- 7:33:46
ChatGPT and Codex Outage and Launch Speculation
ChatGPT and Codex were observed to be down. A possible new-version launch was suggested, but the discussion noted that launches do not normally take a service offline and that something else was likely wrong.
- 7:41:09
Structuring a Review of Pivot Documents
A request was structured to review a folder and understand a pivot, prioritizing the master manifesto before consulting related brainstorming material. The project identity and some details are unclear in the captions.
- 7:44:02
Understanding an Existing Site Before Editing It
The tool is first directed to inspect the site's code and configuration and develop a full understanding of it, as part of an experiment beginning with a simple portfolio-site request.
- 7:47:23
Comparing AI Coding-Tool Behavior
The speaker compares observations about Claude and Codex handling multiple tasks, then describes a separate experience with an unidentified tool where it was unclear whether the requested work had finished after it only opened a port. These are practical impressions, not a controlled test.
- 7:51:29
A Phased Landing-Page Refactor Plan
A prompt is developed to refactor a landing page around a new project's manifesto while retaining its video background and font, with implementation planned step by step and phase by phase instead of as a one-shot change.
Got a question about something in this stream?
Ask in the group