Edition No. 11 · 5 Sep 2026
Twelve repositories for taking money on your own terms
Payments, licences, invoices, and the parts of billing nobody wants to build twice.
Ten editions have been about getting somebody to the door. Measure them when they arrive, translate the page they land on, speed it up, check its grammar, watch it for errors, back it up, follow up by email. A hundred and seventy-seven repositories, and not one of them about the moment a visitor reaches for a card.
That is a strange gap for a SaaS. Pricing is not an afterthought you bolt on once traffic exists — it is a positioning statement, and a product with no way to pay is a marketing problem, because every visitor who would have paid is lost silently and permanently. The same is true of the smaller gates on either side of that moment: the password field that costs you a signup, the consent banner that becomes legally necessary the day analytics goes live, and the visitors who cannot use the page at all.
Eight entries are marketing-shaped, three sharpen Grasppy or ScalpingMate, one is the bench. All twelve are new to the ledger; no repeats today. Three are hidden gems under 3,000 stars. Five further candidates were verified and deliberately left out, named in the appendix with the reason.
This edition, in numbers
| Repos | 12 |
| Hidden gems (<3k stars) | 3 |
| Marketing / product / wildcard | 8 / 3 / 1 |
| New to the ledger | 12 |
| Repeats | 0 |
| Dropped on check | 5 |
Two things worth knowing, separate from the recommendations
- Stirling-PDF is no longer plainly MIT. The root LICENSE is still the MIT text, but it now carves out app/proprietary/, app/saas/, engine/ and several front-end directories under separate licences. Running the free self-hosted build is fine. Assuming the whole repository is MIT is not. - SQLAdmin has changed hands. aminalaee/sqladmin now resolves to smithyhq/sqladmin, and the documentation moved with it. Same shape as the Vale redirect in Edition 10 — old links still work, but the canonical name is different, which is exactly the sort of thing that makes an install guide confusing a year later.
If you only do three things
- Install PostHog Cloud (#8) — this evening. Umami has been on this radar since 21 August and is still not installed. Part of the reason may be that you were told to install four separate things: analytics, session replay, feature flags, surveys. PostHog is all four behind one script tag, and the free cloud tier needs no server at all. Take the cloud tier, not the self-host.
- Put a Polar checkout (#2) behind the pricing page — this weekend. Polar is the merchant of record, which means EU VAT and sales tax are legally theirs to collect and file, not yours. That single fact is what turns "pricing: coming soon" into a business, and it is why this is a weekend rather than a quarter.
- Add CookieConsent (#5) in the same session as PostHog — thirty minutes. The moment an analytics script goes live on a site with European visitors, consent stops being optional. Do it in the same sitting or you carry the exposure for however many weeks it takes you to come back to it.
Every link in one place
| # | Repository | Official site | Stars | Licence | Latest |
|---|---|---|---|---|---|
| 1 | getlago/lago | getlago.com | 10.3k | AGPL-3.0 | v1.52.1 · 27 Aug 2026 |
| 2 | polarsource/polar | polar.sh | 10.2k | Apache-2.0 | daily · 5 Sep 2026 |
| 3 | openmeterio/openmeter 💎 | openmeter.io | 2.2k | Apache-2.0 | 1.0.0-beta.232 · 4 Aug 2026 |
| 4 | logto-io/logto | logto.io | 12.1k | MPL-2.0 | v1.40.1 · 29 May 2026 |
| 5 | orestbida/cookieconsent | cookieconsent.orestbida.com | 5.7k | MIT | v3.1.0 · 4 Feb 2026 |
| 6 | pa11y/pa11y | pa11y.org | 4.5k | LGPL-3.0-only | v9.1.1 · 26 Feb 2026 |
| 7 | scalar/scalar | scalar.com | 15.7k | MIT | release · 13 Aug 2026 |
| 8 | PostHog/posthog | posthog.com | 37.5k | MIT + proprietary ee/ | continuous · Jul 2026 tags |
| 9 | maizzle/framework 💎 | maizzle.com | 1.6k | MIT | v6.0.1 · 16 Jun 2026 |
| 10 | tradingview/lightweight-charts | tradingview.com/lightweight-charts | 17.2k | Apache-2.0 | v5.2.0 · 24 Apr 2026 |
| 11 | smithyhq/sqladmin 💎 | smithyhq.github.io/sqladmin | 2.8k | BSD-3-Clause | 0.31.0 · 6 Aug 2026 |
| 12 | Stirling-Tools/Stirling-PDF | stirling.com | 90.9k | MIT + carve-outs | v2.12.0 · 9 Jun 2026 |
The till
Twelve repositories, checked and reviewed. Every version verified against the GitHub API and dated.
10.5k stars · AGPL-3.0 · v1.52.1 (2026-08-27)
Open-source metering and usage-based billing API — counting what someone used and turning it into a number they owe.
▶Repo detailsthe review · specs · pros & cons · install
What it is
An open-source metering and billing API: it ingests usage events, aggregates them, applies your pricing plan and produces invoices. It covers subscriptions, prepaid wallets, credits, coupons and revenue analytics, and hands the actual card charge off to Stripe or another processor.
Why it matters
Grasppy's real cost is per import — model calls, embeddings, storage — so a flat monthly price either loses money on heavy users or overcharges light ones, and you cannot tell which until you can meter. Lago is what lets you charge per map, per document or per thousand messages without hand-writing invoicing logic, prorations and tax lines. It also solves the thing that actually sinks usage-based pricing in practice: showing a customer, mid-month, what they have spent so far, before the invoice surprises them. And it is the one entry here that lets you change your pricing model later without a data migration, which for a product still finding its shape matters more than the billing does.
- Genuinely usage-based — metering, aggregation, wallets and credits — rather than subscriptions with a usage bolt-on.
- Ships roughly weekly; v1.52.1 landed on 27 August, nine days before this report.
- API-first, so FastAPI talks to it directly and there is no SDK lock-in.
- AGPL-3.0: fine to run for your own business, but read it before linking anything into a distributed product.
- The production topology is Rails plus Postgres, Redis, Kafka and ClickHouse. This is not a small container on a shared box.
- Usage-based pricing is a product decision, not a plugin. Getting the metering model wrong is expensive to unwind once customers are on it.
FastAPI posts a usage event when a Grasppy import completes; Lago aggregates and invoices; Stripe takes the money.
10.2k stars · Apache-2.0 · sdk/1.0.0-alpha.20 (2026-09-04)
Merchant-of-record checkout for software products — the EU VAT problem legally stops being yours.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A billing and checkout platform for software products, open source under Apache-2.0 and run as a hosted service. As merchant of record, Polar is legally the seller: it collects and remits sales tax and EU VAT on your behalf, issues the invoice, and pays you out.
Why it matters
Selling a subscription across borders is not a payments problem, it is a tax-registration problem, and it is the single most common reason a solo founder's paid tier sits at "coming soon" for a year. Polar removes it: you get a checkout link, a customer portal, licence keys, usage-based add-ons and webhooks, and somebody else files the returns. For Grasppy that is the difference between a working Pro tier this weekend and a research project you keep deferring. Set beside Lago (#1) the division is clean — Polar is how you take money now; Lago is how you meter it later, if and when usage-based pricing is worth building.
- Merchant of record: the tax and compliance burden genuinely is not yours, which no self-hosted option can offer.
- Apache-2.0 and shipping constantly — production deploy tags dated today, SDK
1.0.0-alpha.20yesterday. - Checkout, customer portal, licence keys and webhooks work out of the box, with no server to run.
- The code is open; the value is the hosted service. This is a vendor relationship with a percentage attached, not a self-host.
- The SDK is still on
1.0.0-alphatags — expect churn at the integration layer. - A merchant of record owns the payment relationship. Moving away later means re-subscribing every customer by hand.
Set beside Lago (#1) the division is clean — Polar is how you take money now; Lago is how you meter it later, if and when usage-based pricing is worth building.
A checkout link on the Astro pricing page (Ed. 3 #4); a webhook endpoint in FastAPI that flips the account to Pro; the same webhook into n8n to create the Twenty contact and start the Mautic (Ed. 10 #6) welcome sequence.
2.2k stars · Apache-2.0 · v1.0.0-beta.232 (2026-08-04)
Real-time usage metering with entitlements and balances — built for products whose unit of cost is a model call.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A real-time usage metering service aimed at AI, API and infrastructure products: it collects usage events, aggregates them continuously, and enforces entitlements and balances. It sits between your application and whatever eventually bills the customer.
Why it matters
Grasppy's marginal cost is a model call, so the moment you offer a free tier you must answer "has this account used up its allowance?" on every single request — cheaply, without scanning a table. That is what an entitlement is, and it is the piece people hand-roll badly and then discover is wrong at the worst moment. OpenMeter is also the cleanest way to put a live "62 of 100 maps used" bar into the product, which is a retention and upgrade prompt as much as an accounting detail. It overlaps Lago (#1) directly: OpenMeter meters and gates, Lago meters and invoices. Read both and pick one — two counters that disagree is worse than neither.
- Built for token- and request-shaped usage, including entitlements and balances, not just end-of-month invoicing.
- Apache-2.0, so it can live inside a paid feature — unlike Lago's AGPL.
- Active and small: beta.232 on 4 August, with releases most months this year.
- Still
1.0.0-betaafter 232 tags — the project is not yet promising API stability. - The production deployment wants ClickHouse and Kafka; the light footprint is the development one.
- 2.2k stars and a young company behind it. This is a bet on a vendor as much as a dependency.
The remaining balance reads straight into the React header; the totals feed Lago or Polar for the actual charge.
A FastAPI dependency that checks the entitlement before an import starts and posts the usage event when it finishes.
14.5k stars · MPL-2.0 · v1.43.0 (2026-08-31)
Identity service with social sign-in — because 'choose a password' is the most expensive field on the internet.
▶Repo detailsthe review · specs · pros & cons · install
What it is
An open-source identity service: hosted sign-in pages, OIDC and OAuth 2.1, social and enterprise connectors, multi-factor, role-based access and multi-tenancy. Your app redirects to it and gets a token back.
Why it matters
Every field between a YouTube viewer and their first Grasppy map is a place people quietly leave, and "choose a password" is the most expensive field on the internet. Social sign-in is the cheapest conversion win available to you right now, and Logto is one of the few options that ships connectors your Russian-speaking half actually uses — Yandex and VK alongside Google, GitHub and Apple — which is the practical form of the second-audience argument Edition 4 made. It also takes off your hands the part of authentication you should least want to own as a non-programmer: password resets, session revocation, brute-force protection, and the security bugs hiding in all three. MPL-2.0 is file-level copyleft, the mildest of the copyleft licences to live next to a commercial product.
- Social connectors covering both halves of a bilingual audience, not just the US providers.
- MPL-2.0 is markedly easier to live with commercially than the AGPL options elsewhere on this page.
- Self-hostable, with a managed cloud tier if you would rather not run another service.
- Replacing authentication in a live product is a migration, not an install. Do this before you have paying users, not after.
- A Node/TypeScript service with its own Postgres database — another container competing for the Hetzner box.
- Last release 29 May, about three months ago. Healthy, but not the weekly cadence of Lago or Scalar.
MPL-2.0 is file-level copyleft, the mildest of the copyleft licences to live next to a commercial product.
The React shell redirects to Logto; FastAPI validates the JWT on every protected route.
5.7k stars · MIT · v3.1.0 (2025-02-04)
Vanilla-JS GDPR consent with real script blocking — the banner that becomes necessary the day analytics goes live.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A small, dependency-free consent plugin written in plain JavaScript, with categories, per-service toggles, translations and a documented API for blocking scripts until the visitor agrees. Around 30 kB, no framework and no build step required.
Why it matters
PostHog (#8), Umami and any pixel you ever add all write to storage, and the moment one of them goes live on a site with European or UK visitors you need consent captured before the script runs — not a decorative banner that appears after the tracker has already fired, which is what most implementations actually do. The ordering genuinely matters here, which is why it is one of today's three things: install it in the same sitting as your analytics or you carry the exposure for however many weeks it takes you to come back. The bilingual detail is real too — it handles translations natively, so a Russian visitor sees a Russian banner rather than an English one, which is the same argument Edition 4 made about everything else on the site.
- Vanilla JavaScript — it drops into the Astro site and the React app equally, with nothing to compile.
- Real script-blocking rather than a cosmetic banner, which is the half most implementations get wrong.
- Multi-language support built in, so EN and RU are configuration rather than a fork.
- Seven months since v3.1.0. It reads as finished rather than abandoned, but it is not fast-moving.
- It gives you the mechanism, not the legal advice — which categories you need is still your call, and this report is not a lawyer.
- The default styling is generic; making it look like your site is a CSS afternoon.
A script tag on the Astro site and inside the React shell.
4.5k stars · LGPL-3.0 · 10.0.0 (2026-08-28)
Command-line accessibility auditing that names the failing element, not just the score.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A command-line accessibility testing tool: it loads a page in headless Chrome, runs it through the HTML CodeSniffer and axe rule sets, and reports the WCAG failures with the exact element responsible. A companion CI mode checks a whole list of URLs in one run.
Why it matters
The European Accessibility Act has applied to consumer-facing digital services since mid-2025, which puts a bilingual, EU-facing SaaS squarely in scope — this is a compliance surface, not a nice-to-have. Beyond that, a Grasppy map that cannot be reached with a keyboard is a defect you will never hear about, because the affected visitor simply leaves; the react-arborist tree from Edition 7 #11 exists precisely because the canvas is unusable for a whole class of person. Unlighthouse (Ed. 2 #9) already gives you a Lighthouse accessibility score, but a score is not a task list. Pa11y names the failing element on each page, which is the difference between a number you feel bad about and a morning's work you can actually finish.
- LGPL-3.0-only. Run it as a tool; do not link it into anything you ship.
- Automated checks catch roughly a third of real accessibility problems. Colour contrast judgement and focus order still need a human pass.
- Six months since v9.1.1, with a small maintainer team. Healthy, but not a fast-moving project.
Standalone tool — run it from the Mac against the live site, or run pa11y-ci over the Astro sitemap in the deploy step so a regression is caught before it ships.
16.1k stars · MIT · release-2026-09-07-42fc8b1 (2026-09-07)
Turns the OpenAPI file FastAPI already generates into documentation a person would read.
▶Repo detailsthe review · specs · pros & cons · install
What it is
An open-source API platform: a reference renderer that turns an OpenAPI or Swagger document into a browsable, searchable documentation site, plus a built-in client for trying requests on the page. It is a drop-in replacement for the default Swagger UI that FastAPI ships.
Why it matters
FastAPI hands you a complete OpenAPI schema for free and then renders it with a page nobody enjoys looking at — which starts to matter the moment Grasppy has any surface facing another developer: an integration, a webhook, a connector, or simply the API you show when selling to a team. Well-rendered API docs are also content in the plainest sense: they are indexable pages, they answer long-tail queries, and they are exactly the structured, quotable material that geo-optimizer-skill (Ed. 8 #3) and llms-txt (Ed. 2 #7) want an answer engine to find. And it is nearly free — you are replacing one line of FastAPI configuration, which makes it the cheapest item on this page after the consent banner.
- Near-zero effort on your stack: point it at the
openapi.jsonyou already serve. - MIT, and releasing constantly — the tags are literally dated, most recently 13 August.
- The embedded client lets a reader try a call without leaving the documentation.
- Only earns its keep if Grasppy has an API somebody outside your head will use. If it does not yet, this is a bookmark.
- A large, fast-moving TypeScript monorepo — pin a version rather than tracking a CDN's latest.
- Beautiful documentation will not rescue a badly shaped API. This is presentation, not review.
Swap FastAPI's built-in docs handler for the Scalar renderer over the same /openapi.json , then publish it as a route on the Astro site (Ed. 3 #4) so the pages are crawlable rather than hidden behind the app.
39.6k stars · MIT, with a proprietary ee/ directory · desktop-v0.61.263 (2026-09-08)
Analytics, session replay, feature flags, experiments and surveys behind one snippet — four standing recommendations collapsed into one.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A product analytics platform that bundles event analytics, session replay, feature flags, experiments, surveys and error tracking behind a single snippet. It is fully self-hostable, but in practice it is a cloud product with a large free monthly allowance.
Why it matters
This ledger has recommended Umami for analytics (Ed. 1, repeated in Ed. 5), GrowthBook for flags and experiments (Ed. 3 #6), OpenReplay for session replay (Ed. 9 #3) and Formbricks for surveys — four installs, four databases, four things to back up. Fifteen days after the first of them, none exists. PostHog is all four behind one script tag, and the honest recommendation is the cloud free tier rather than the self-host: you get an answer to "why do people leave the map screen?" this evening instead of after a ClickHouse deployment. The trade deserves stating plainly — you are sending visitor behaviour to a US company instead of keeping it on your own box, which is exactly the privacy position Umami and remark42 (Ed. 5 #1) were chosen for, and exactly why CookieConsent (#5) sits on this page beside it.
- Replaces four separate standing recommendations with one snippet. This is a decision as much as a tool.
- Generous free cloud tier — at your current traffic the bill is plausibly zero for a long time.
- MIT core and genuinely open, so moving it onto your own hardware later is a real option rather than a marketing claim.
- Self-hosting is a serious undertaking — ClickHouse, Kafka, Postgres, Redis. Edition 10's warning against standing up OpenReplay and Langfuse in one month applies here twice over.
- The
ee/directory is proprietary, so "open source" carries an asterisk. - Cloud means visitor data leaves your server, which contradicts the privacy-first story the rest of your stack tells. Decide which story you are telling, and tell it consistently.
The trade deserves stating plainly — you are sending visitor behaviour to a US company instead of keeping it on your own box, which is exactly the privacy position Umami and remark42 (Ed.
5 #1) were chosen for, and exactly why CookieConsent (#5) sits on this page beside it.
One script tag in the React shell and on the Astro site, loaded from CookieConsent's analytics category.
1.6k stars · MIT · v6.1.3 (2026-09-07)
Write email in utility classes; it compiles to the inlined table markup Outlook still requires.
▶Repo detailsthe review · specs · pros & cons · install
What it is
An email development framework: you write HTML email with utility CSS and reusable components, and it compiles to inlined, table-based markup that survives Outlook and Gmail. It is a build tool that produces files, not an editor with a canvas.
Why it matters
Both email options in this ledger — Listmonk (Ed. 2 #1) and Mautic (Ed. 10 #6) — expect you to supply the HTML, and Edition 6 answered that with GrapesJS, a drag-and-drop builder. Maizzle is the other answer, and it fits how you actually work far better: campaigns become text files, which means Claude Code writes your five-email onboarding sequence and Git remembers every revision, instead of you dragging blocks around a browser and hoping you can reproduce it. The bilingual case closes the argument — the Russian edition of that sequence is the same templates with a different strings file, not five emails rebuilt by hand. Paired with Mailpit (Ed. 6 #3) you can see precisely what lands, in every client, before a subscriber does.
- Text in, HTML out — so Claude Code can write and revise campaigns and Git can track them.
- Solves the genuinely hard part: CSS inlining, table layout and Outlook's conditional comments.
- MIT, small surface area, and v6.0.1 landed in June.
- A Node build step, and a second toolchain sitting beside Astro's.
- No visual editor at all. If you want to drag blocks, this is the wrong tool and GrapesJS is the right one — pick one and delete the other.
- 1.6k stars and a very small team. A niche tool rather than a platform.
Build the templates locally, then paste or API-push the compiled HTML into Listmonk or Mautic.
17.2k stars · Apache-2.0 · v5.2.1 (2026-08-12)
TradingView's own charting library — the chart a forex trader already knows how to read.
▶Repo detailsthe review · specs · pros & cons · install
What it is
TradingView's own open-source financial charting library: candlesticks, lines, areas and histograms drawn on an HTML5 canvas, around 45 kB, built for long series and smooth panning. It draws price data. It does not compute indicators.
Why it matters
Edition 10 gave ScalpingMate stockstats (#9) — fifty legible indicators under a licence you can ship — and then had nowhere to draw them. This is that screen. Everything else in ScalpingMate's chain (tsfresh, Optuna, arch, SHAP) is research machinery a paying user never sees; the chart is the product, and adopting TradingView's visual conventions means a forex trader recognises it in the first second instead of learning yours. Apache-2.0 matters here for the same reason it did for quantstats (Ed. 4 #11): this can live inside the paid screen, which nothing AGPL in your backtesting stack can.
- Apache-2.0 and from TradingView itself, so the conventions are the ones your users already have in their hands.
- Very small and fast — years of M1 bars without the page dying.
- Canvas-based and touch-friendly, so it survives the phones your YouTube audience arrives on.
- Deliberately minimal: no indicator maths, no drawing tools, no order overlay. You build all of that.
- v5 was a breaking rewrite, so most tutorials and Stack Overflow answers still target v4 and will mislead you.
- Their terms require an attribution notice on the chart. Read the licence page before shipping a paid screen.
A React component in the ScalpingMate front end, fed by a FastAPI endpoint that returns stockstats columns as series.
2.8k stars · BSD-3-Clause · 0.31.1 (2026-09-01)
A SQLAlchemy admin panel mounted inside the FastAPI app he already runs — support without a psql prompt.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A SQLAlchemy admin interface for FastAPI and Starlette: you declare which models to expose and get a browsable, searchable, editable web UI over your tables. It mounts as a route inside the application you already run — no separate service, no second database connection.
Why it matters
You are one person supporting two products, and every support email — "my import failed", "did my payment go through?" — currently ends with you at a psql prompt reconstructing a query. This is an afternoon of work that turns those into a search box, and for someone who does not write SQL from memory that is the difference between answering in a minute and putting it off until Sunday. It also makes the rest of this ledger inspectable rather than theoretical: Procrastinate's job table (Ed. 5 #10), Lago's usage events (#1), Logto's users (#4) all become something you can look at without a terminal. Note the move: the project is now smithyhq/sqladmin, and it is releasing roughly monthly under its new home.
- Exactly your stack — FastAPI, Starlette, SQLAlchemy — with nothing new to deploy or monitor.
- BSD-3-Clause, the most permissive licence on this page.
- Actively released under the new owner: 0.31.0 on 6 August, after 0.30, 0.29 and 0.28 in the three months before.
- An admin panel over live production data is a serious authentication problem. Put it behind Logto (#4) or a VPN — never a shared password.
- Still 0.x, and the ownership change means older tutorials point at the previous repository.
- It exposes your schema exactly as it is. A confusingly named column stays confusingly named in the UI.
Mounted at /admin on the existing FastAPI app, reading the same SQLAlchemy models.
91.5k stars · MIT at the root, with app/proprietary/, app/saas/, engine/ and several frontend directories carved out under separate licences · v2.14.3 (2026-08-06)
About fifty PDF operations in one self-hosted container — nothing leaves the machine.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A self-hosted web application offering around fifty PDF operations — split, merge, compress, convert, OCR, redact, sign, fill forms, rotate — through a browser interface, running entirely on your own hardware. One Docker container, no upload to anybody else.
Why it matters
Two threads in this ledger end in PDFs. Typst (Ed. 8 #7) produces the bilingual lead magnet you will trade for an email address, and it will want compressing before it goes on the site. And Grasppy's whole document-ingest question — Docling (Ed. 3 #9) against Marker (Ed. 9 #9), which Edition 9 asked you to settle on ten real PDFs — needs a corpus of genuinely messy multi-page files, which means splitting, rotating and OCR-ing documents you already have. Doing either on a free upload-your-file website means posting unreleased marketing material and other people's documents to a company you know nothing about. This is the wildcard for a reason: it earns nothing this month, and then one afternoon it saves you.
- Around fifty operations in one container, entirely offline — nothing leaves the machine.
- Very actively released; v2.12.0 in June brought large memory improvements to merging.
- Useful across both products and the video workflow, not just one corner of one of them.
- The licence is no longer simply MIT — see the note above. The root is MIT;
app/proprietary/,app/saas/andengine/are not. - A JVM application that wants real memory headroom. Read Beszel (Ed. 5 #8) before adding it to the Hetzner box.
- Never expose it to the internet unauthenticated. It will happily process whatever anyone uploads, at your expense.
# 1 — PostHog. Nothing to install at all. # Create a free project at https://posthog.com, copy the snippet it # shows you, and load it from CookieConsent's analytics callback # rather than pasting it directly into the page head. # 2 — Polar. Also nothing to install. # Create an account at https://polar.sh, add a product and a price, # copy the checkout link onto the pricing page, then add one webhook # route in FastAPI to hear about payments. SDK docs: https://docs.polar.sh # 3 — CookieConsent, in the same sitting as PostHog. npm install vanilla-cookieconsent # No build step? Use the CDN module documented at # https://cookieconsent.orestbida.com
Checked, and left out
These were opened for this edition and did not make it, with the reason.
juspay/hyperswitch
juspay/hyperswitch (43.0k, Apache-2.0, v1.123.1 2026-05-19) — verified alive; dropped as far too heavy for a founder with no checkout at all. Backlogged.
killbill/killbill
killbill/killbill (5.7k, Apache-2.0) — no readable release date and a Java/Tomcat deployment; Lago covers it with a provable cadence. Backlogged.
usesend/useSend
usesend/useSend (4.7k, AGPL-3.0) — REDIRECT from unsend-dev/unsend. No releases at all, recency unprovable. Backlogged; it fills a real hole and is worth re-checking.
Rikorose/DeepFilterNet
Rikorose/DeepFilterNet (4.6k, MIT/Apache-2.0) — no dated releases, recency unprovable. Was the intended YouTube slot. Backlogged.
google/schema-dts
google/schema-dts (1.2k, Apache-2.0, v2.0.0 2026-03-23) — verified healthy and a real gem; dropped only for slate fit. Backlogged and flagged as a strong candidate for an early edition. PROMOTED AND RECOMMENDED THE NEXT DAY as Edition 12 #9.