Fieldwatch: a ham radio blogger turned a "Flock detector" into an Android app, and one Spanish tweet brought 1,053 stars in a day

Fieldwatch is an Android app that only listens, never transmits: your phone already picks up nearby Wi-Fi and Bluetooth broadcasts, and it matches them against 250 device signatures to tell you whether there are Flock license-plate cameras, police body cams, Meta glasses, drones, or trackers nearby. The library was created in Sep 15, and by Nov it had 6 4,342 stars and 516 forks. Below: what it is, why this one took off, where the stars came from, who it's competing with, and what to copy and what to avoid if you want a project that gets stars too.

One tweet: "This guy putting stuff on GitHub shouldn't be free"

On Sep 30, 2026, in the afternoon (US Eastern time), Florida ham radio operator and off-grid communications creator Off Grid Pete posted on X: "Fieldwatch 1.1.16 is out!" It was one more routine update in a half-month stretch, and the repo was gaining only a few dozen stars a day at the time.

Six hours later (UTC 23:49), @CopyRebeldia, a Spanish-language account that specializes in reposting "GitHub gems," posted:

Lo que este tipo subió a GitHub no debería ser gratis. (This guy putting stuff on GitHub shouldn't be free.) An app that turns your phone into a surveillance detector. It listens to the Wi-Fi and Bluetooth signals around you and tells you what's nearby: hidden trackers, police body cameras, drones and their pilots' locations, Meta glasses that are recording, and gunshot-detection microphones. It recognizes 242 kinds of devices and will warn you if one is following you. No account, no server, no network, open source.

This account's previous post was "an engineer at a small workshop in Morocco uploaded a complete phased-array radar to GitHub, for free." Same template, different project. The next day (Oct 1), Fieldwatch gained 1,053 stars; Oct 2 645, and Oct 3 299 (daily star counts from trendshift).

This tweet is basically the app's feature list copy-pasted. Fieldwatch got reposted not because the tech is new, but because its feature list is a good post by itself.

What it actually does

The author says in the README that this is a tool he built for himself: "I wanted to see which Wi-Fi hotspots and Bluetooth devices my phone can pick up, to figure out what's running around me." It's entirely passive. No external hardware, no account, no backend, works offline.

• Device recognition: 250 signature sets sorted by category. The Flock camera set, for example, matches specific hardware address prefixes and Wi-Fi names starting with "Flock-" (MakeUseOf, 10-22). You can add your own signatures, or flag a category to alert whenever it shows up.
• Tracker alerts: flags devices that follow you around. "Hunt" mode uses signal strength to walk you step by step toward a specific Bluetooth device.
• Anomaly warnings: possible GPS jamming, your phone connecting to a cell tower it has never used in this location, or a sudden burst of Bluetooth pairing requests.
• Maps and reports: plots the devices you heard along your route. A stakeout session generates a report you can compare against later.

Not on the Play Store. Install the APK from GitHub via sideloading; the app checks GitHub for updates on its own. Reached v1.1.42 in 52 days, and at the end of October one release added 27 languages.

Why now

2026 in 9, Flock Safety became one of the hottest public issues in US tech. The New Yorker wrote that it wants "a world under tight surveillance with no way out" (9-09), Wired reported hackers breaking into Flock cameras (9-16), and TechCrunch covered a federal judge calling it "indiscriminate mass surveillance" (10-03). Each of these hit 500-600+ points on Hacker News. BBC reported people smashing Flock cameras in various places, and someone in Congress proposed the Ban Flock Act. A lot of ordinary people are asking one question: is there one near me?

Before this, "Flock detectors" were a thing for ESP32 hobbyists: the open-source firmware flock-you (1,501 stars) required buying a board and flashing it. Off Grid Pete also made an ESP32 version (GoFlockYourself, 67 stars). Fieldwatch moved the same signatures onto Android phones, which everyone already carries. MakeUseOf's headline was literally "no extra hardware needed."

Where the stars came from

(Daily star counts from trendshift. The most recent stretch uses GitHub's daily numbers, as of 11-06.)

First, the author brought a small audience of his own. Off Grid Pete has about 18k followers on Instagram and kept posting Reels and X updates. That audience wasn't enough to push the project to thousands of stars (it added only a few hundred in the second half of September), but it kept supplying material: screenshots, short videos, one-line feature descriptions that repost accounts could use directly.

Second, a consumer repost chain, not a developer community. After the Spanish post, reposts clustered from Oct 1 to 3: Instagram's @techstartup.ai, Spanish and English shares on LinkedIn, privacy communities on Lemmy, and repost sites in several languages. Articles followed from Cybernews (10-02, headline: "This free app can detect Flock cameras, creepy glasses and other trackers"), Reclaim The Net (10-09), Hackers-Arise (10-13) and MakeUseOf (10-22).

What didn't work: someone submitted it to Hacker News on Oct 3, and it got 2 points. It has never appeared on GitHub's official trending page. This is an open-source project whose stars came from ordinary people sharing it, not from developers.

Downloads say more than stars. Across 29 releases, the APKs have been downloaded 27,523 times in total (GitHub, 11-06). The latest three releases each have a couple hundred downloads, not counting direct links from the README. A real group of people installed it and keeps following the updates.

Who it's competing with

This category was crowded in 2026. Fieldwatch is the one that is "on your phone, checks everything, and is packaged best."

The ones worth watching are RFSentinel and Flock-You-Android. They do nearly the same thing, one earlier and one with more features, and both stalled at a hundred-odd stars. What they lacked wasn't code. Nobody wrote the "this shouldn't be free" post for them.

What it still gets wrong

• Accuracy is a problem. ALPR Maps, which sells a Flock map app, published an article stating flatly that "no phone app can detect an unmapped license plate camera by listening to radio." These cameras mostly upload over cellular. Bluetooth or Wi-Fi is used only during installation and setup, and is often turned off once they're running. Passing at 70 km/h, you spend less than a second in Bluetooth range. The article also cites a similar app that flagged a treadmill (Bowflex T6) as a camera. It has a commercial stake, but the physical limits it describes are real. Fieldwatch's own disclaimer also says it can produce false positives and miss real devices.
• Android only, and it has to be sideloaded. iPhone users can't use it. A regular person who sees a repost and wants to try it first has to enable "install unknown apps."
• Interest is fading. Daily adds have dropped from over a thousand to twenty or thirty, and the Flock topic itself will pass.

If it were you

The short verdict: building another "check for Flock or surveillance with your phone" app will most likely produce a third RFSentinel. What's worth learning is the approach: when a public issue peaks, turn something that used to need hardware and technical skill into a version an ordinary person can run on a phone, and make its features easy to list one by one.

Who this suits: developers who know a bit about radio, hardware or Android internals, who are already part of an interest community (ham radio, off-grid, privacy, outdoors, cars), and who have a small audience and can keep posting short-video updates.

Who it doesn't suit: people who want to ride the same topic for quick hype; people who expect to post on HN or GitHub and wait for developers to star it (that path did nothing this time).

How to do it:

1. Look for the combination of "a topic at its peak + something that used to require buying hardware." When the next hot topic appears, ask yourself: can the thing hobbyists build with ESP32, SDR or a Raspberry Pi be turned into a good-enough version that runs on a phone alone?
2. Write the features as a list where every item is something an ordinary person understands at once and finds a little alarming or satisfying ("alerts you if a device is following you"). A repost account can copy-paste that into a post.
3. Your posture has to be credible: passive reception only, offline, no account, open source, false positives stated clearly. For privacy tools, users judge whether you can be trusted before they look at features.
4. If you do build in this category, go after what Fieldwatch left undone: an open, field-tested signature library (which signature was confirmed in the wild, and what the false-positive rate is), or a version for surveillance devices in other countries (UK ANPR, European enforcement equipment, for example), where the audience and topic are out of Fieldwatch's reach. On iPhone the system doesn't let apps scan Wi-Fi, and Bluetooth is restricted too, so confirm the thing can actually be built before you start.

How to get stars: post a phone recording of the app with every release on X and Instagram so repost accounts have material to pull from. After launch, actively find "GitHub tool" accounts in Spanish, Portuguese and Chinese; this time the spark was a single Spanish account. Attach the APK to GitHub Releases and build in auto-update so people who installed it stay.

When it's already too late: the Flock window has basically closed, with only tens of new stars a day now. The approach itself doesn't expire. Wait for the few weeks when the next "mass surveillance anxiety" or similar topic appears.

Signs you've misjudged it: when you read your own feature list, you don't want to share it. Or within two weeks of launch no repost account or media outlet has picked it up on its own, and stars stay in double digits. That means the topic isn't hot enough or the list isn't "instantly understandable." Changing the angle does more good than posting the same thing again.