funnyfoobar
Joined 3 years ago
Joined 3 years ago
Some of these are present here: https://github.com/vamsipavanmahesh/claude-skills/
Planning to package this as a workshop, so companies could be benefit from AI Native SDLC.
Put together the site yesterday https://getainative.com
Couple of the people I have worked with in the past agreed to meet me for a coffee, will pitch this. Fingers crossed.
I am not talking about vibe coding here, which is totally different and understandable. But I am talking about the professional context where the structure is already in place.
Here is my workflow, if it helps:
1. I have a system level prompt i execute(command/skill) before I start anything which says things like
i) follow SOLID principles
ii) follow TDD
iii) follow these testing principles like, don't test the implementation details etc
and 20 other things
2. I give it the context of what I need to accomplish, and toggle the plan mode ask it follow the above things I have set at system level, and generate me a document to review step by step3. I review everything, add feedback to the generated plan, prompt again, and finally finalize the plan
4. now that the plan is finalized, I would have a small alignment with co-workers and ask it implement step by step, and then before proceeding with next step, i will tell it to ask me for feedback
each steps code is reviewed and commited, before proceeding to next step
5. once all steps are done, manual testing is done, i do a local review
6. I have a specific skill which reviews everything, and gives me feedback
7. we have human reviewers who will review the code, and coderabbit
8. then we take it to the staging and uat before taking to prod
so see we have not skipped any process of "software engineering", claude code is just an accelrant
but if you are not doing most of the above steps, i see what you are saying would happen by default
Things have changed drastically now, engineers with these tools(like claude code) have become unstoppable.
Atleast for me, I have been able to contribute to the codebases i was unfamiliar with, even with different tech stacks. No, I am not talking about generating ai slop, but I have been enabled to write principal engineer level code unlike before.
So i don't agree with the above statement, it's actually generating real value and I have become valuable because of the tools available to me.
the thing about being an engineer at commercial capacity is "maintaining/enhancing an existing program/software system that has been developed over years by multiple people(including those who already left) and do it in a way that does not cause any outages/bugs/break existing functionality.
while the blog post mentions about the ability of using AI to generate new applications, but it does not talk about maintaining one over a longer period of time. for that, you would need real users, real constraints, and real feature requests which preferably pay you so you can priortize them.
I would love to see such blog posts where for example, a PM is able to add features for a period of one month without breaking the production, but it would be a very costly experiment.
> AI can write 15k lines of code, but it cannot take Liability for a single one.
Thanks for writing this, I needed it.
But my question is "how many of those will be needed", because I am not saying that programmers are not needed.
When less numbers are needed, there will be so much competition in finding those jobs, esentially would also mean not able to find the work, as there will be always someone who would be willing to the job at lower wage and come to work with more youthful energy.
Just speaking out loud.
If everyone is doing high level stuff like architecture and design, how many of "those people" will be really needed in the long term? My intuition is telling me the size of market needing number of engineers will shrink.
I kept seeing the same thing across engineering teams: everyone bought Copilot or Cursor seats, people used it for a few weeks, and then output didn't really change.
The tooling is good but teams are treating AI like autocomplete instead of integrating it into how they actually ship code.
My take is that AI starts making a real difference when you apply the same discipline you already have for software engineering.
Plan the work, implement, write tests, commit, open a PR, get it reviewed. If you just vibe code with no structure around it, you get messy diffs and hallucinated tests that nobody trusts.
So I'm building around that workflow. Think of it as giving AI the guardrails of your existing eng process.