Knowledge Base

Building High-Quality Inputs

Keywords – Track trends in your target niche (for example Canadian shopping or new game launches) and craft keywords around them. Feed those into the Keyword Scraper to expand the list.
Parameters – Run sample dorks with the correct ext: filter (such as ext:php) and harvest repeated parameters from the top results.
Dorks – Strong keywords plus relevant parameters and the right page types form the foundation of high-performing dorks.

Restart the client if the module refuses to launch.
Ensure at least one engine is selected. Proxyless plans should run Google only without proxies; other plans need proxies when scraping Google.
Match threads to your hardware (50–75 on 4 vCPU / 8 GB RAM, 100–200 on 8 vCPU / 16 GB RAM).
Disable Windows Defender or other security software that might block traffic.
Test with and without a VPN if you use one.

Confirm the correct machine is selected and proxies are working. Try a quick run without proxies to compare.
Use thread pairs suited to your hardware (50/50 main-secondary for 4–6 vCPU, scale up for stronger machines).
Enable the Auto-dumper feature and configure presets before starting.
"No match" indicates the presets did not match any columns; refine them or run the DB Explorer to inspect the structure first.

Balance the requested dork count with the volume of keywords and parameters. Generating millions of dorks from a small input set will stall.

The Dehasher matches hashes against known data; it does not brute-force them. If a hash stays unresolved, move on to the next line or use external cracking tools.

Google scraping is exclusive to proxyless plans. Upgrade from the dashboard (Upgrade button) if you need it. You'll pay the difference based on your current plan.
Enable Keep unfiltered in Scraper settings if you specifically want raw URLs without parameters.

Low DPM: check proxy quality, network stability, thread values, and the quality of your dorks.
Low FPM: review dork quality and thread counts, and remember that the local AntiPublic system filters duplicates.
Expect roughly 5% injection and 5–10% dump rates. Load enough links to hit your targets, then refine presets to increase yield.

Client will not open – Check the machine limit (max three). Install Microsoft WebView2 if missing. Pause all tasks; if the issue persists, delete problematic tasks.
Client shows zero values – Confirm you have not hit the machine limit, disable antivirus, re-enter the API key, and test network access (AnyDesk is a good benchmark).
Client keeps crashing – Stop running tasks one by one to find the culprit. If necessary, delete and recreate machines.
Machine not listed – Restart the client and re-add the API key. If the network blocks outbound connections, switch to another server.

Downloads remove bad or empty rows (email: or :pass). Ensure timeouts and threads are tuned so the dumper captures complete rows.
If storage is full without uploads, clear old runs in History.
When saving Dumper results hangs, try an incognito browser session. If that fails, reset the task with fewer threads.

Renew – Dashboard homepage → Add days.
Upgrade – Dashboard homepage → Upgrade.
Restore presets – Reset the Dumper from the settings page.
Export injectables – Use the ... menu next to Save results to export injection successes before rerunning with different presets.

Deleting domains.db removes the AntiPublic cache if you intentionally want previously scraped URLs.
The Dork Checker uses different APIs than the Scraper; if you trust your dorks, run the Scraper directly.
Idle status in the Dumper means injection succeeded but presets did not match. Adjust presets or minimum row filters.
Use the blog post at https://www.stationx.net/google-dorks-cheat-sheet/ for additional dork construction techniques.