Meen el hodhod dah?
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 practical stream centered on an unfinished “Mission Control” tool for visually supervising coding-agent swarms, exposing their status and consumption, and detecting stuck loops, alongside live reviews of the interface, bird characters, and animation. The discussion broadened to goal-based roadmaps, separating software-building tools from company-operations systems, and balancing custom agent harnesses against added complexity and waste. Luqman answered audience questions about Codex, Claude, AI tools, learning, fundamentals, and hosting, repeatedly framing many comparisons as personal experience rather than conclusive tests. He also argued that distribution and demand validation can be harder than technical construction, and discussed a possible open-source GitHub release with community contributions or a competition whose prize and judging details remained undecided. The interface, motion, roles, naming, and release plan were still incomplete, and neither real token savings nor practical success had yet been established.
141 moments from the stream
- 1:04:06
Focus or Run Multiple Projects in Parallel?
The discussion suggests that AI tools have made parallel project work easier, while distraction remains the central constraint. The speaker also argues that a strong idea alone does not attract an audience; reach and distribution must be solved separately.
- 1:05:25
Will AI Reduce the Value of SaaS?
The speaker offers a cautious opinion that advancing AI may reduce the value and production cost of many SaaS products even when underlying demand remains. He argues that businesses ultimately want outcomes such as stronger sales or visibility, and future AI systems may create the needed tools for a goal instead of requiring a fixed software purchase.
- 1:08:36
Study the Project-Building Tools, Not Only the Project
The speaker redirects attention toward the tools and methods used to build and operate projects, presenting them as more useful learning material than the identity of any single project.
- 1:09:33
Check Demand and Distribution Before Blaming the Product
The speaker encourages persistence when a builder genuinely believes people need the product, arguing that weak results can reflect a reach and distribution problem rather than a product defect. He qualifies this by noting that some products lack real demand, so validating the need remains essential.
- 1:14:18
Do You Need More Tools Beyond Claude Code and Codex?
Answering an audience question, the speaker argues that Claude Code and Codex have become largely self-contained and usually do not require a large supporting toolset, while acknowledging that some skills can still help when needed.
- 1:15:32
Build Your Own Tools for Large-Project Management
The speaker recommends identifying a project's needs first—such as a roadmap, an overall view, and a solid foundation—then using Codex or Claude to build a tailored command center or tracking system instead of relying on a few initial prompts. He also says he intends to develop the idea seriously and open-source it for community contributions.
- 1:17:21
An Idea Alone Does Not Determine a Project’s Success
The speaker argues that a struggling venture does not necessarily have a bad idea; execution and the operators’ experience may be decisive. He suggests that the same idea could produce a different outcome in more experienced hands, presenting this as an opinion rather than a guaranteed rule.
- 1:21:36
From Individual Tasks to Missions Run by Agent Swarms
He explains his expectation that development will become more autonomous, with related tasks grouped into a single “mission” handled by multiple agents. He argues that a command center showing agent status and allowing an orchestrator to intervene when work stalls may save time and tokens, while warning that swarms can consume resources unnoticed without monitoring.
- 1:24:58
Distinguishing Project-Building Tools from Management Systems
He distinguishes Mission Control, intended to build a new project or company through work missions, from another system for recurring operations across areas such as customer service, debugging, and marketing. He recommends maintaining one command center suited to the workflow rather than creating one per project, and choosing the system according to whether the work is constructive or continuously recurring.
- 1:30:29
Does the Project Actually Need a VPS?
He questions adopting a VPS merely because creator recommendations have become popular and urges evaluating the project’s actual needs first. He explicitly avoids assuming the questioner followed the same pattern and allows that a VPS may be appropriate depending on the requirements.
- 1:33:13
A Personal Preference Among AI Coding Tools
He says he tried alternative models at the audience’s request but did not find them suitable for his workflow. As a personal preference, he recommends Codex within the choice being discussed while acknowledging that Claude may offer slightly better quality in some respects; this is experiential opinion, not a documented benchmark.
- 1:35:12
Experimenting with Constant Background Sound for Focus
He describes an experimental background track that differs from dub techno and uses a constant sound or selected frequency with the aim of supporting focus. He presents it as an ongoing experiment rather than claiming proven effectiveness.
- 1:36:26
Reading Hosting Recommendations Through Affiliate Incentives
He discusses how recommendations for hosting and commerce platforms may be tied to affiliate programs and audience conversion through intermediary pages. He does not say a sponsored opinion is worthless, but argues that viewers should recognize the commercial incentive because the promoted product may not be the best fit for everyone.
- 1:42:09
Redesigning the Thinking Symbol
A visual review argues that the thinking gear looks like a generic tool symbol and asks for a more expressive representation of thought.
- 1:43:21
First Impressions and the Value of Information
The speaker says he downloaded an unnamed tool but did not explore it deeply, while finding its self-improving-agent idea compelling. He then discusses how sharing information can bring non-monetary returns as well, framing this as a personal view rather than an absolute rule.
- 1:44:34
Comparing GPT, Claude, and Gemini for Arabic Copywriting
The speaker considers GPT, Claude, and Gemini good copywriting options, but personally prefers Gemini and GPT for Arabic and finds Claude comparatively mixed; this is an opinion, not a systematic test.
- 1:46:47
The Hoopoe and Agent Swarms
The speaker explains that the hoopoe is a character in a tool he is building to work more effectively with agent swarms and the wider project. He imagines several birds with different roles, led by the hoopoe, and explores an open-source direction after weighing an existing system against building a graph-based automation system.
- 1:51:27
Plan to Release the Tool as Open Source
The speaker said that, at the time of recording, he expected to release the tool as open source after preparing a domain and website, welcome contributors, and possibly create a discussion space if interest grew.
- 1:52:02
Ducking Background Music During Speech
After audience feedback that the music was too loud, the speaker discussed lowering it and using an effect that automatically ducks it while he speaks, while preferring a calm background that helps him focus.
- 1:54:24
A Visual Command Center for Coding Agents
The concept visualizes agent sessions, their status, current work, and interactions with code, GitHub, and testing, alongside token consumption and usage-limit tracking. The broader aim is a project command center, with stuck-loop detection potentially helping avoid unnecessary consumption.
Got a question about something in this stream?
Ask in the group