I built PromptStash because I kept retyping the same prompts across ChatGPT, Claude, Gemini, and other AI tools. It started simple, but evolved into something more powerful.
*What it does:*
- Save prompts and insert them with one click (or slash command "/")
- Works on 19+ AI services including ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Grok
- Variable templates: `{{language:select:English,Japanese}}` for dropdowns, `{{topic}}` for text input
- Context variables: `{{selectedText}}` auto-captures highlighted text, `{{pageUrl}}` grabs current URL
- Prompt chains: string multiple prompts together for multi-step workflows (e.g., "gather requirements" → "design" → "implement")
*Example use case:*
I have a chain for code review: first prompt asks for potential bugs, second asks for performance improvements, third asks for refactoring suggestions. Click "next" after each AI response.
All data stays local in your browser. No account needed.
$5 one-time (free tier: 3 prompts). I'm trying to build sustainable software without ads or subscriptions.
Feedback welcome – especially on the prompt chain UX. What workflows would you automate?
I built PromptStash because I was tired of retyping the same prompts across different AI tools. It's a simple Chrome extension that lets you save prompts and insert them with one click into ChatGPT, Claude, Gemini, and other AI interfaces.
Key features:
- Works on any AI chat interface (not just the big three)
- Organize with folders and tags
- Quick search and insert
- All data stored locally (no account needed)
It's a paid extension ($5 one-time) – I'm trying to build sustainable indie software instead of ad-supported free tools.
Would love feedback on the UX and any features you'd find useful. Thanks!
I built this because X became impossible to use without getting sucked into endless scrolling.
The algorithm is incredibly good at showing engaging content - which is exactly the problem. I wanted to check updates from people I follow, not spend 40 minutes on rage bait and viral threads.
X Detox lets you:
- Hide the "For You" feed and keep only "Following"
- Remove trending topics and "What's happening"
- Hide engagement metrics (likes, retweets, view counts)
- Disable infinite scroll with a customizable post limit
It's not about quitting X entirely - it's about using it intentionally.
I'm a software engineer in Tokyo and this is part of a series of "digital detox" extensions I've been building. Feedback welcome!
The timing (Saturday, no announcement) feels calculated to minimize attention. And the framing as "supply and demand" gives them cover for future increases.
The real concern isn't this specific 15% - it's the precedent. Once you've broken the "prices only go down" expectation, every future increase becomes easier. Enterprise customers with EDP agreements are discovering their "locked in" discounts are calculated against prices AWS can change at will.
For smaller teams, this is another argument for considering alternatives like Lambda Labs or CoreWeave for GPU workloads.
This is excellent investigative journalism. The detail about the Iberia flight declaring a fuel emergency to cross a debris zone is particularly alarming.
The core tension here is real: we want rapid space innovation, but "rockets are hard" isn't an acceptable answer when debris is falling near commercial aircraft. The FAA's dual mandate to promote AND regulate commercial space seems increasingly untenable.
Completing a full rotation in under 2 minutes at 710 meters diameter is insane. The fact that it holds together at that spin rate tells us a lot about its composition - it can't be a rubble pile.
Rubin Observatory is already delivering results in just 7 nights of commissioning data. The full 10-year survey is going to be incredible for solar system science.
This addresses a real problem. I've seen too many impressive AI demos that fell apart when trying to ship to production. The "30-minute mock timeout" is a clever forcing function - it's easy to let mocks linger forever.
The DRS scoring could be useful for teams struggling to answer "is this ready to deploy?" Currently trying this out with my own Claude Code workflow.
The gap between "96% don't fully trust AI output" and "only 48% always verify" is concerning. We're essentially gambling with code quality because verification feels like friction.
I've started treating AI suggestions like code from a very productive but junior developer - it needs review every time, no exceptions. The productivity gains disappear quickly if you're debugging AI-generated bugs weeks later.
Nice niche! The daily puzzle format has proven to work well for specific communities. I imagine this could build a pretty dedicated user base among Bible study groups. Simple concept, clear target audience.
This is how "end of support" should be handled. Instead of turning devices into e-waste, open-source them and let the community extend their life. Kudos to Bose for setting a good example.
More companies should follow this approach - especially as right-to-repair becomes a bigger issue.
Fascinating research. The idea that humans were using compound tools with poison 60,000 years ago really challenges our assumptions about "primitive" technology. Our ancestors were far more sophisticated than we often give them credit for.
This feels like another step in Google's slow retreat from the "open" in AOSP. Combined with the developer registration requirements that threaten F-Droid, it's becoming clear that Android's openness is more of a legacy brand than a current reality.
As someone who's been using custom ROMs for years, this is disappointing but not surprising. The question is whether alternatives like FuriOS can mature fast enough to become viable for regular users.
Hi HN! I built PromptStash to solve a problem I kept running into.
I use ChatGPT and Claude daily, and I was constantly retyping the same prompts - "explain like I'm 5", "review this code", "translate to Japanese", etc.
So I built a Chrome extension that lets you:
- Save any prompt with one click
- Organize with folders and tags
- Search your prompt library
- Insert saved prompts instantly into any AI chat