Building $1M ARR in SaaS 🚀 | AI agents that work while you sleep 🤖 | Shipping daily, failing forward | Documenting the my ups and down so you don't have to 💪
Remote
Joined June 2015
- Tweets8.2K
- Following456
- Followers783
- Likes12.6K
Pinned Tweet
I got tired of manually copying env vars between 8 different AI coding tools and researching for hours for the exact config format then end up messing everything else.
🚨 Hence, I built Confkit — Never spend a minute on MCP config again!
✅ Auto-sync all MCP clients (codex, cc, gemini etc)
✅ One click MCP addition to all environments
✅ Profile switching for handling multiple env with unique MCPs/envs
No more copy-pasting API keys at 3am.
#buildinpublic #indiehacker #DevTools
There's a day comes in your adult life when you think okay maybe this is not for me.
I don't want to miss out on these fun holidays, family gatherings, date nights and maybe I should have taken that job which paid well.
These days are just a test of how badly you want it and are willing to put your head down, ignore all the clout, get the users and win.
Weekly limit got reset few last night
Cloudflare drops vulnerability skill
Lets try it, what's the worst that can happen
Wake up to 50% usage on my $200 weekly limit (started from 10%)
And it was only halfway done.
The skill found some really good concerns though but only if i had more usage to actually fix them :(
With the recent wave of vulnerability disclosures, it’s a good time to take a closer look at your own code.
Turn your coding agent into a team of security auditors with the open-source skill that seeded Cloudflare’s vulnerability discovery harness.
The 6 phase approach:
1. Map the codebase
2. Hunt for vulnerabilities
3. Challenge each candidate
4. Record the findings
5. Verify the findings
6. Report what’s confirmed, rejected, or still needs validation
Start your hunt here:
github.com/cloudflare/securi…
The fact that claude has shipped gigantically in a lot of horizontal niches impacting more and more % of your daily lives, while building better models every day.
Is an example to the pace with which RSI (self-feedback loops) is coming along , one day you can blink a startup out of existence just because it's a personal need of yours.
It's hard to imagine needing something and not having it in a couple years of time.
These day's I just feel bad for gemini man, it's been trying to force it's way into AI in so many places.
I like using it as an assistant but damn I hate all my buttons now show "would you like AI", god no!
.@ClaudeDevs can we have a way to start a task chip agent with a particular model + effort setting, rather than the one which I'm using in the current window right now?
I like to tweak them based on the task so it'd be a great addition to the app!
The kind of codebases I'm working on right now, which were written by humans pre-GPT era, for corporate internal apps.
Sometimes it feels like I wish AI had written that code, it'd atleast have a single line of documentation or pattern to what it's writing.
We can all shout AI makes mistakes, but I think in a few years AI would be solving problems with code that's beyond human understanding.
Being so busy and self dependent with things that you love doing with the people you love,
that you forget to post about it on social media and live everyday as if it's your last, is my /goal state. :)
I wanted AI to automate my cooking, cleaning, laundry so I could do more creative work and work on fun projects.
Now AI has automated my creative work, so I'm doing cooking, cleaning, laundry myself. 🤦♂️
The more I grow in my freelance/client facing consultatnt career, the more I realise 👇
Clients don't actually know what they want themselves.
They'll demand you things left and right.
But it is our duty to consult them to gather all information to know their true intentions/problems, then find all calculated solutions and present it to them to pick.
And without a doubt, AI is so helpful at doing this. We can ship hundreds of presentations a day and consult the client the best way to go.
But giving exactly how the client wants without knowing their true intentions is definitely going to get you kicked in future.
Going to bed while giving tasks to cloud agents and waking up and getting them merged one by one.
This literally feels outsourcing your work to freelancers at fraction of cost and without the mind-numbing experience.
Well you can even ask playwright in cloud to keep your test screenshots ready and babysit the PRs for sometime to get all reviews fixed before you spending anytime on it.
Future of development is going to be even more awesome!
Damn it, @DecartAI the future of meetings and live streaming are going to be insane! 🔥
Tried out Lucy 2.5 the new realtime video editing from your live cam itself and it's scary good.
Thank you so much Kimi K3 and GPT Sol, it wouldn't have been possible without you!
🖤
Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits.
Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit.
Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity.
Anyone using claude code cloud agents still?
I used to love it but they only have these three edit modes, and having "auto" mode blocks half of my legitimate requests to run necessary actions.
@ClaudeDevs
Thanks google gemini code assist.
You've had a good run.
Curious, what are y'all migrating towards right now?
I'm using codex heavily on PR reviews but they do hit rate limits (even for PR reviews) every now and then.
Here's how you setup Kimi K3 to be used with @opencode 🖤
Don't forget to add the image input as modalities, or else it won't be able to see image inputs. (see 2nd screenshot)
There is: docs.orcarouter.ai/integrati…
