Code Driven Manifesto
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 an unfinished hands-on review of Burnish and its Code Driven Manifesto workflow: generating, editing, and exporting a manifesto; turning it into internal instructions for the storefront, articles, visibility, reputation, and discovery; and considering how its status and Autopilot integration should appear. Testing surfaced issues with generation, progress, navigation, state updates, scoring, auto-publishing, and background jobs. Some next-batch logic appeared to work in one attempt, but critical bugs and questions about cancellation, timeouts, concurrency, and persistence remained under investigation. Luqman emphasized backend wiring, user experience, and debugging before interface polish, arguing that premium-positioned software needs commensurate reliability while acknowledging that no launch can eliminate every bug. Other discussions covered an exploratory Telegram community workspace, audience questions about learning programming with AI and the advantage of experience, uncertain opinions about coding-tool quality, pricing, and context capacity, and acquisition and retention measurement without invented figures, plus shorter conversations about YouTube music copyright, Algerian music, personal goals, and screen-use glasses.
85 moments from the stream
- 4:06:45
Cancelling Hung Jobs and Scoping Follow-up Work
The discussion notes that a genuine abort signal should prevent a job from leaving a worker hanging. It also asks for a plain-language explanation of a follow-up flag and whether the improvement belongs in the current pull request or a later one.
- 4:08:00
Reviewing the Visibility Page’s First-Use Experience
A user-perspective review asks what someone sees after installing the app and first landing on the visibility page. It emphasizes a direct, concise explanation and considers how any needed improvement affects pull-request scope.
- 4:09:11
Starting a New Phase with Supabase
The work clarifies prerequisites and the next step for beginning a new development phase that depends on Supabase, followed by monitoring a merged change as it moves into deployment. Private account and project details are excluded.
- 4:11:10
Preserving Work Context Before Context Compaction
The discussion instructs the assistant to retain the system-prompt wiring, the brainstorm outputs, and the next follow-up task before context compaction so implementation can continue coherently.
- 4:15:06
Defining the Next Phase After Applying Changes
After processing and applying the changes, the discussion asks what phase comes next and what users should see, then states that a change was merged and deployed successfully before moving to a follow-up branch or fix.
- 4:20:17
Prioritizing Product Tabs by Impact and Actionability
The discussion simplifies a five-tab product and proposes that Reputation and Discovery are the highest-impact tabs because they relate to competitor citations and responses; they are also described as the only actionable tabs. The next step requested is logical discussion rather than a sketch.
- 4:22:02
Upgrading Reputation and Discovery After Manifesto Creation
A proposal for generating an internal system prompt from the Manifesto to guide brand voice and influence several areas of Burnish. Before it exists, Reputation and Discovery would show basic questions while encouraging its creation; afterward, those questions would be replaced automatically with tailored ones and presented as Manifesto-powered. The discussion focused on communicating this user journey and requested a visual mock-up of its stages, not yet on technical implementation.
- 4:32:38
Reusing the Marketing Skill from Project Documentation
A request to inspect the project documentation, find a previously created marketing skill, and invoke it. The excerpt ends before the desired result is fully stated, so its intended output remains unclear.
- 4:33:00
Updating Unit Economics After a Pricing Change
A discussion of updating average revenue per user and then reassessing downstream metrics such as cost per install, while avoiding assumptions about an uncertain sales tier that could distort the estimates.
- 4:33:52
Showing Whether Content Is Powered by the Manifesto
A proposal to indicate whether questions are powered by the manifesto, later simplified after observing that the state applies to the whole group rather than to each question individually.
- 4:37:32
Why Changes Become Harder After Launch
An explanation that post-launch constraints are not only technical: major changes can affect customer experience and force new decisions to account for what already exists, making a full application refactor possible but difficult.
- 4:40:13
Adapting Manifesto Status Across App Sections
The discussion stresses visually distinguishing an active manifesto from a recommended one, adapting the panel content to each section, and preparing mock-ups for the articles and storefront areas.
- 4:43:33
Timeouts and Concurrency for Background Jobs
Practical questions about an AI service hanging within a background-job queue: how long the timeout should be, what determines the limited execution lanes, and whether increasing concurrency would actually improve performance.
- 4:46:49
Auditing the Acquisition-Cost Model
A critical review of a metric described as ‘cost per trial retained install,’ calling for complete calculations and assumptions, an explanation of the role of average revenue per user, and clarification of cost-per-action versus cost-per-install terminology before relying on the model.
- 4:50:49
What Happens When an AI Agent Misses a Call?
A discussion of system behavior after a missed call: whether it should retry automatically, wait for the next scheduled call, and how that state should be defined clearly.
- 4:52:28
Turning an Idea into a Task in a Separate Session
A workflow suggestion to capture an issue discovered during brainstorming as a task, open a separate session to implement it, and keep the original discussion moving locally.
- 4:57:47
Measuring Direct Ads Through Install and Retention Costs
A discussion of sending ads directly to an install page while recognizing that installs alone are an inadequate measure. The proposed analysis focuses on cost per install and the cost of users who remain active during the trial, using industry benchmarks and sensitivity analysis rather than invented assumptions. It also raises how instances where the AI fails to respond should be counted.
- 5:06:37
Reducing Technical Explanations to Clear Sentences
A request to restate a set of downstream technical-wiring points in plain English, using one sentence per point so they are easier to understand and review.
- 5:09:26
Do Cybersecurity Capabilities Justify Higher Model Prices?
The speaker considers a claim that cybersecurity models are entering an expensive capability race, but remains uncertain whether the cited capabilities or controls justify the higher price, suggesting that operation itself may simply be costly.
- 5:11:03
Replacing Brand Settings with Manifesto Generation
The speaker discusses removing the old brand-identity settings and offering a simpler option to generate a brand manifesto, presented as the proposed basis for store optimization, while cleaning brand text before sending it to the AI.
Got a question about something in this stream?
Ask in the group