// per-site profile
what detects you on bot.sannysoft.com
bot.sannysoft.com is defended by a detection vendor being profiled. below: the layers it enforces and the tool config the matrix predicts survives it.
enforced detection layers
not yet measured: enforced layers on bot.sannysoft.com
this entry is in the catalog but carries no scored capability yet. the matrix updates as the benchmark self-heals, so check back or see the capability matrix.
notes
Intoli / fpscanner headless test table (bot.sannysoft.com). JS appends '(passed)' or '(failed)' to each test row and colours it; a clean browser renders every row '(passed)' with zero 'failed' (VERIFIED live, 0 occurrences on a real Chrome, WebDriver row 'missing'), while any leaked automation tell (webdriver present, empty plugins, permission clash, broken UA) flips a row to '(failed)'. So human = the rendered page carries no 'failed'. Passive FINGERPRINT verdict, no interaction. A stale/early read errs to a conservative 'fail', never a false pass. KAN-165. SIGNALS (WP3.10, verified live 2026-07-24): the first table holds 11 data rows; `tr:has(td)` excludes the TH header cleanly. Each row is captured TWICE on purpose. The cell TEXT is the observed value (`Google Inc. (NVIDIA)`, `5`, `en-US,en`) while the cell CLASS carries the pass/fail (`result passed`). Text alone loses the verdict for exactly the rows that report a value instead of the literal word `passed`.
questions
- what detects you on bot.sannysoft.com?
- bot.sannysoft.com's detection vendor is being profiled.
- which config survives bot.sannysoft.com?
- no enforced layers are scored yet, so no prediction is available.