Android Click-Fraud Test Automation Tool
Budget: €30 – €250 EUR
I need an Android-only automation package that lets my team safely reproduce and measure three specific adversarial patterns—click spamming, ad stacking, and click injection—inside our own sandbox. The solution must run equally well on emulators and on physical-device farms, switching between them with no code changes.
Core idea
You will ship an installable APK paired with repeatable scripts that drive the app, trigger the chosen fraud scenario, collect detailed logs, and reset the device state afterward. All traffic must be hard-coded to our internal mock pages; the tool must never attempt to touch live Google Ads or any production endpoint.
Why it matters
These synthetic attacks help us harden our in-house click-prevention engine, so realism (randomised timings, variable screen sizes, network conditions) is important, yet safety is non-negotiable.
Deliverables
– Signed debug APK ready for side-load
– Automation scripts (Bash, Python, or similar) that launch the app, select a behaviour, and iterate for an adjustable run count
– Read-me covering setup on both emulators and popular cloud device farms
– Structured log output (JSON or CSV) capturing timestamp, behaviour type, coordinates, and device identifier
– A simple config file where we can toggle spamming rate, stack depth, or injection delay without recompiling
Acceptance criteria
1. Running the default script on an Android 13 emulator and on a real Pixel device yields identical logs.
2. No outbound calls are made beyond the whitelisted mock domains we will supply.
3. Each fraud mode exhibits at least a 90 % success rate over 100 iterations without crashing.
Mention any frameworks you plan to use—Appium, ADB, Espresso, etc.—so we can verify compatibility with our lab before awarding the project.
Core idea
You will ship an installable APK paired with repeatable scripts that drive the app, trigger the chosen fraud scenario, collect detailed logs, and reset the device state afterward. All traffic must be hard-coded to our internal mock pages; the tool must never attempt to touch live Google Ads or any production endpoint.
Why it matters
These synthetic attacks help us harden our in-house click-prevention engine, so realism (randomised timings, variable screen sizes, network conditions) is important, yet safety is non-negotiable.
Deliverables
– Signed debug APK ready for side-load
– Automation scripts (Bash, Python, or similar) that launch the app, select a behaviour, and iterate for an adjustable run count
– Read-me covering setup on both emulators and popular cloud device farms
– Structured log output (JSON or CSV) capturing timestamp, behaviour type, coordinates, and device identifier
– A simple config file where we can toggle spamming rate, stack depth, or injection delay without recompiling
Acceptance criteria
1. Running the default script on an Android 13 emulator and on a real Pixel device yields identical logs.
2. No outbound calls are made beyond the whitelisted mock domains we will supply.
3. Each fraud mode exhibits at least a 90 % success rate over 100 iterations without crashing.
Mention any frameworks you plan to use—Appium, ADB, Espresso, etc.—so we can verify compatibility with our lab before awarding the project.