A few weeks ago I shipped my own Chrome extension and hit the wall every indie maker knows: the listing is live, it actually works… and it has zero reviews, so nobody trusts it enough to install it. Reviews are a chicken-and-egg problem, and there's no honest way to solve it alone.

So I built PeerReview — a review exchange for indie developers. It's already live and being used, and we're doing our official Product Hunt launch today - https://www.producthunt.com/products/peerreview?launch=peerreview

How it works:

• List your product (needs a live listing on a supported platform — Chrome Web Store, Product Hunt, Play Store, App Store, etc.)
• Your queue shows the whole open pool of products you can review right now. Pick any one.
• Leave a GENUINE review on the real platform — a real opinion, not a scripted 5-star.
• Paste the link back; the owner reads it and marks it Verified.
• Give one honest review to earn one back. No stockpiling, no bots.

Where it's at so far (all real, before we've even launched on PH):

• ~37 developers in the pool
• 54 products listed
• 100+ genuine reviews exchanged

Two things I want to be upfront about, because I know this space attracts skepticism:

1. This is NOT "buy 5-star ratings." That's against store policy and, frankly, useless — fake praise doesn't convert and doesn't help you improve. The value is real feedback from people who actually build things. A thoughtful 3-star is worth more than a fake 5.
2. Reviews live on the real platforms, not on PeerReview, because that's where they're actually useful to future users. PeerReview just facilitates the exchange and verifies it happened.

It's free (every feature on every plan), privacy-first, no trackers — it's built under the tapdot umbrella where that's a hard rule.

You don't have to wait for the launch — it's live now if you've got a product that's under-reviewed: https://peerreview.tapdot.org

I'd genuinely love your honest feedback on the tool itself. What would make this actually useful for your next launch?

Spent the last few weeks building Quill, a Chrome extension that runs AI writing tools (rewrite tone, proofread, summarize, compose) entirely on-device using Chrome's built-in Gemini Nano model — no cloud API, no accounts, nothing leaves the browser.

Got into this because every AI writing tool I'd used had the same catch: to fix a sentence, it had to leave your machine first. On-device inference removes that step entirely — there's no server to log, breach, or retrain on your drafts, because there's no request at all.

Tradeoffs I Ran Into Building It:

• On-device models are weaker than cloud LLMs on anything needing broad knowledge or long reasoning — fine for transformation (tone, grammar, summarizing), not for research-heavy generation.
• It's still gated behind Chrome flags and device requirements (~4GB VRAM), so there's real first-run friction before someone hits the "aha" moment.
• Free tier: 3 tones unlimited + 5 uses/hour on the utility tools. Pro unlocks 23 tones + custom tones + unlimited usage.

Launching on Product Hunt July 14th: https://www.producthunt.com/products/quill-15

Would love feedback — especially on whether the free/Pro split feels fair, and whether the privacy angle actually matters to people vs. just being a "nice to have."

Been spending the last few weeks building on top of Chrome's built-in on-device AI (Gemini Nano, via the Prompt/Writer/Rewriter/Summarizer APIs) instead of calling out to a cloud LLM, for a writing-tools extension called Quill (rewrite tone, proofread, summarize, compose — all in-page).

A few things I ran into that might be useful if you're building something similar:

1. Availability is inconsistent. Each dedicated API (Rewriter, Writer, Summarizer, Proofreader) can be enabled/disabled independently of the others, and independently of the general-purpose Prompt API. I ended up building a capability map per install (ai_caps) so the UI only shows/enables what's actually usable on that machine, with the Prompt API as a fallback whenever a dedicated engine isn't available.
2. The flags UX is rough for end users. Everything's still behind chrome://flags/#prompt-api-for-gemini-nano and friends — a content script can't navigate a page there, so I route it through the background service worker to open the flags page for the user. Even then, there's real first-run friction before someone gets to the "aha" moment.
3. Model download + first run. First use downloads the model (multi-GB), so I pre-trigger the download on install rather than making the user hit it cold on their first real action.
4. The upside: no server, no accounts, nothing the extension sends off the device — genuinely different privacy story than any cloud-API-based extension, since there's no request to intercept or log in the first place.

Curious if others building Chrome extensions on these APIs have hit similar issues, especially around fallback behavior when an engine isn't enabled.

The extension just got approved, and for more details, please refer to https://tapdot.org/quill/.

Drop your Chrome extension below and I'll genuinely install it, use it, and leave an honest review on the Web Store. Real feedback, not a drive-by 5 stars. If something's broken or confusing I'll say so (nicely).

Here's why I care about this: I shipped my own extension a while back and hit the wall every indie dev hits — you can build something good, but with zero reviews on the listing, nobody trusts it enough to try it. I started leaving honest reviews on other people's extensions in threads like these, and a few of them did the same for mine. Not inflated ratings — actual "here's what I liked, here's what annoyed me" reviews. It genuinely helped, both the products and just knowing another dev had looked at my thing.

The problem is these threads are chaotic and one-off. You post, you get three reviews if you're lucky, it scrolls away, done.

So I built PeerReview → https://peerreview.tapdot.org/ (completely free).

It's a small web app that matches indie devs to give each other genuine reviews. You list your product, you get matched with someone else's, you leave a real review on the actual platform (Web Store, Product Hunt, etc. — not on my site), you paste the link back, and the owner reads it and marks it verified. Everyone reviews everyone else's stuff over time instead of one dev shouting into one thread.

A few things I want to be upfront about, because this space attracts shady "boost your ratings" junk and I'm not that:

- It's about genuine feedback, not gaming stars. Reviews are honest and can be critical. If all you want is fake 5-stars, this isn't it (and that violates Google's policy anyway).

- Free. Actually free. All features on every tier — you only pay if you want to list a ton of products. It's under tapdot, which is not-for-profit.

- Privacy-first. No trackers, no third-party data selling. Google sign-in is only used for account identity. Your name/email/photo are hidden on your public profile unless you turn sharing on yourself.

- It's very early — I'm literally launching it to this sub first because this community is where the idea came from.

I'd love for you to try it and tell me what's broken or dumb about it (its still in BETA - I started with this community so wanted to share here before launching elsewhere). And genuinely — drop your extension in the comments regardless and I'll review it, tool or no tool.

Cheers 🙏

CommentIQ reads the comments on any YouTube video and gives you:

• Sentiment breakdown — what % are positive / neutral / negative
• Top topics — what themes keep coming up
• Audience questions — pulled out as a list so you don't have to scroll to find them
• Content ideas — things your audience is suggesting or asking for

The whole thing runs on-device using Chrome's built-in AI (Gemini Nano). Nothing is sent to any server. It works on your own videos and on competitors' videos — useful for both.

It's free to install: https://chromewebstore.google.com/detail/commentiq/cmkehjfedgcjnpjojlanlgalojnlcmjf

Would love feedback from creators who actually use it — especially on whether the topic clustering is surfacing the right things.

Found this snake skin in our apartment parking(Madipakkam, Chennai), Can anyone help identify the snake, will it be a recurring visitor. And how do I proceed further in informing the right official?

Hey all — I've been using Chrome's side panel API for a personal project and ended up building something I actually use daily, so figured I'd share it here for feedback.

It's called Margin.

The idea: instead of a news feed you scroll forever, it's a stack of cards in the side panel — one headline + short summary at a time, swipe (or scroll) to the next. Pick a handful of topics once, tap through your day's news in a couple of minutes, tap any card to read the full thing at the source. No infinite scroll, no algorithm trying to keep you there.

A few things I cared about building it:

- No account, no tracking, no analytics — it's local-only, settings and reading stats never leave your device

- No backend — it fetches RSS straight from publishers (BBC, The Guardian, NYT, etc.) on your machine

- Optional on-device AI summaries via Chrome's built-in Summarizer, so nothing gets sent to a server

- 25 topics to choose from, a few themes (light/sepia/dark/etc.)

It's free to use with 3 topics; there's an optional $1.99/mo tier for unlimited topics + extra themes + a reading-insights dashboard, just so it's not totally unsustainable to maintain. Being upfront about that since I know paid tiers can be a turnoff — wanted to mention it before anyone asks.

Download link: https://chromewebstore.google.com/detail/kmiddbhoklpkheeopgoehknnkejohnhp?utm_source=item-share-cb

Would genuinely love feedback — especially on the reading flow, whether the topic selection feels right, or anything that feels broken. This is a side project, not a startup, so blunt feedback is welcome.