I stopped asking my AI what to build — now it just knows what task I mean (CLI experiment)
I’ve been experimenting with a CLI workflow that removes the most annoying part of AI-assisted dev:
re-explaining context every single time.
Instead of prompts like:
“Integrate Stripe payment gateway with X, Y, Z…”
I just run:
blocpad task 1
And it already knows:
the full task description
related docs & flows
pricing, decisions, constraints
comments from PM / lead / previous discussions
No re-prompting. No context loss.
What changed for me:
Tasks feel stateful, not conversational
AI behaves more like a teammate than a chatbot
CLI → instant context → real work
It’s early, rough, and very opinionated — but it made me realize how broken “prompt-based” workflows actually are.
Curious:
Do you prefer chat-first AI or task-first AI when building real products?
36 views
Replies
This is a really interesting take on the "context problem" with AI tools. I work with Claude Code daily and the constant re-explaining is definitely the biggest friction point - especially when you come back to a project after a few days and have to rebuild the whole mental model for the AI again.
BlocPad - Project & Team Workspace
@virtualviki 100% — the problem isn’t the model, it’s context reset.
Blocpad keeps context in tasks and state, not chat, so you don’t have to re-explain everything after a break.
Where does this hurt you most with Claude Code?
@mihir_kanzariya The biggest pain for me is cross-platform context. I'm building an app for both iOS and Android simultaneously, and Claude loses track of which platform I'm working in - I end up re-explaining platform-specific constraints every session. Sometimes I literally work in two terminals at the same time just to keep both contexts alive.
BlocPad - Project & Team Workspace
@virtualviki This is a perfect example of context living in the wrong place.
Platform shouldn’t be something the model “remembers” — it should be part of the task’s state. Running two terminals just to keep iOS/Android straight is basically compensating for missing structure.