Sable Project

@hovinthenorth Yes I have Premium + and was using it yesterday. Spent about 2 hrs having it build an iPhone app with pretty good results.

Grok Build open to Premium+ subscribers? I just got it and I don't have Grok Heavy:

@Bhavani_00007 There will still be sooo many people that don’t know how, or don’t want to build their own products. They also are buying SaaS for the support that comes with it. There are lots of things in business that have gotten easier but people still pay for it.

Everyone says: "SaaS is dead" "Software is free everywhere" Cool!!!!! Then go build your own Slack, GitHub, Notion, Zoom, 1Password, Jira, and Rippling this weekend. What's stopping you? Seriously, I want to know.

@TTrimoreau Coming from the beer world, almost every craft brewery made an IPA. Does that mean you can’t open with an IPA? Only a small few are actually building their own apps. Despite what your X timeline is telling you.

If everyone is building with the same AI tools… what makes your product worth choosing?

@sflorimm cd ~/

What command do you use most on the terminal?

@huge_icons sableproject.com. Working on two apps right now but have a couple tech reviews up already.

Drop your project URL

@LucaCaponeX I am working on two apps right now. My biggest struggle is whether I put them on the App Store for free, $0.99 one time, a yearly charge, or use advertisements. What is your plan?

I am very excited about the mobile app that I'm building. It's going to be my first iOS mobile app. And this is what it does. Have a look below. I made this video with Claude Code. What does it look like?

Will report back with the scars and wins but so far I was able to open the app on my iPhone. So that’s a start! #VibeCoding #AppleApps #RoadTrip

What I’m actually tracking: • How much planning happens before it starts generating • Where the output feels production read vs a nice demo • How much of my job as PM still has to happen vs. the model just “getting it” and what I need to add to my initial prompts.

Tried something new this weekend: asked Grok to one shot an entire app build. Not “help me with this function”. Straight “here’s the idea, go build it.” Curious to see how far these things can actually carry a non dev PM before it needs adult supervision.

@IBuzovskyi I haven’t within Hermes. I’m just using it from Ghostty. I am downloading Hermes right now.

@Sable_Project That`s a good point. Thank you . Also Grok Build is awesome paiting with Hermes Agent. Have you tried it ? x.com/IBuzovskyi/sta…

A software engineer with 10 years of experience says he builds entire side projects from his phone using Claude Code without reading a single line of code. → plan mode only. read the plan. read it again. if something’s unclear, stop and ask before a single line gets written. → if the plan doesn’t fit in your head, it’s too big. cut it in half. → argue with the agent during planning. this is where the real work happens. everything after is just execution. → version control before anything runs. if it breaks you want a rollback, not a rewrite. → have the agent write test cases in plain english. not code. plain english. so you can actually read what it’s supposed to do. → only after all of that: hit auto mode. people debate whether vibe coding works. it works when 80% of your effort goes into thinking and 20% into letting the AI execute. the prompt is the last step. the plan is the product. which of these are you skipping?

@ryancarson That’s been my main goal the past couple months. Have agents “do my job” but be really good and building the systems that give them the ability to do those tasks. Pretty much making SOPs for agents.

I'm becoming convinced there is going to be an explosion of jobs for people who are great at using agents. There's a clear line I'm seeing between people: 1. Use Claude Code / Codex / Devin / etc 2. Use ChatGPT / Gemini / Perplexity Distinction is simple: If your agent writes persistent code, you're in group 1 and the future is v v bright.

Making lots of changes on how we write Jira tickets and how we breakdown epics. Heavier on the User Stories and adding the Don’t Wants to the ticket. Moving to writing tickets for Agents not Humans.

None of this is settled. The map isn’t written yet. That’s the exciting (and slightly terrifying) part. PMs and operators navigating this, what are you questioning hardest right now? #ProductManagement #Roadmapping #AIAgents #NowNextLater #ProdMgmt

I’ve also been questioning: • Features: Does this still need a full spec or can we build, test, get user feedback faster? • PRDs: Are we writing for humans or agents? • Dev meetings/cadence: Sprints still make sense? • Estimation: T-shirt sizing by tokens & ease of UAT?

I’ve been thinking about my role during this AI transformation for awhile and have many questions: 1) Do multi quarter calendar roadmaps still make sense when AI agents are compressing timelines from months to days/weeks? (among other things)

@sabir_huss50540 Do you think Product Roadmaps over a time frame (ex: next two quarters) is the best way to do things anymore? I’m looking to move to a Now, Next, Later and ditching the calendar style Roadmap.

PROMPT 3: The Product Roadmap Builder You are a VP of Product who has built and defended roadmaps at high-growth startups and scaled enterprises, earning trust from both engineering and executive teams. Build a complete product roadmap for [PRODUCT NAME] covering the next [TIMEFRAME]. Context: Current product state: [DESCRIBE] Strategic goals for this period: [LIST] Known user problems to solve: [LIST] Engineering capacity: [TEAM SIZE AND VELOCITY IF KNOWN] Stakeholder constraints: [DEADLINES, COMMITMENTS, OR DEPENDENCIES] Biggest prioritization challenge: [DESCRIBE] Deliverables: ① PRIORITIZATION FRAMEWORK ‣ Scoring criteria for this product at this stage (impact, confidence, effort, strategic fit) ‣ How to weight each criterion for your specific context ‣ How to handle items that score well but feel wrong ‣ How to say no to good ideas without destroying relationships ② INITIATIVE BREAKDOWN For each major initiative: Problem statement (whose problem and why it matters) Success metric (how you know it worked) Effort estimate (small/medium/large with reasoning) Strategic fit (which goal this serves) Dependencies and risks ③ SEQUENCING LOGIC ‣ Why this order and not another order ‣ Technical dependencies that constrain sequencing ‣ Market timing considerations ‣ How to build learning into the sequence (test before full build) ④ ROADMAP COMMUNICATION ‣ Executive version (one page, outcome focused) ‣ Engineering version (initiative level with dependencies) ‣ Customer-facing version (what to share and what to protect) ‣ How to handle roadmap questions you cannot answer honestly yet ⑤ ROADMAP GOVERNANCE ‣ How often to review and update the roadmap ‣ What triggers an unplanned roadmap change ‣ How to handle urgent requests without blowing up the plan ‣ How to communicate changes without losing stakeholder trust Build a roadmap that is honest, defensible, and actually followed.

🚨BREAKING: Claude can now think like a top Silicon Valley product leader (for free). Here are 10 Claude prompts that build complete product strategies, user research frameworks & roadmaps in 4 hours: (Product managers are already saving this)

@predotdev I have not found this issue in two of our products. But we have a clean code base and good documentation. I work in ERPs, and we tell our users “Shit data in, shit data out”. I feel like the same goes for coding agents. Messy starting code base = messy code output.

Why do AI coding agents completely fall apart the second you drop them into an existing codebase? It isn't a context window issue. It's something much more fundamental...