What is Context Aware Detection?
Using surrounding words to decide if a match is a real leak.
Short answer
Context aware detection examines the words around a potential match to decide whether it is actually sensitive. A ten character token in a normal sentence is different from the same token next to the words access key, and context signals are what tells them apart.
In depth
Plain regex matching has a false positive problem: random strings, order numbers, and sample data look like credentials. LeakSnitch boosts or penalizes scores based on context signals like sensitive keys, disclosure phrases, and question patterns, so real leaks score high and innocent text passes.
Entropy analysis works as a fallback when context is missing, catching obfuscated secrets by their randomness. The combination is why LeakSnitch can claim near zero false positives without hiding its rules.
Frequently asked questions
How does context reduce false positives?
Context signals add to the detection score when surrounding text suggests a credential, and dummy and placeholder word lists subtract from it. The scoring is transparent and published in the pattern config.
Related terms
Stop AI leaks on the device, not in the cloud.
Install LeakSnitch in under 30 seconds and protect every prompt sent to ChatGPT, Claude, Gemini, and 25+ AI tools. Free for individuals.