fingerprint-optimizer

Automatically evaluates and optimizes engine configurations to minimize Fingerprint JS suspect scores and eliminate bot flags.

SKILL.md

GitHub

Fingerprint Optimizer Skill

This skill empowers Hermes to automatically heal fingerprint decay. If anti-bot mechanisms evolve and the Golden Template begins failing, Hermes can trigger this skill to empirically determine a new optimal configuration.

How it works

When invoked, the Fingerprint Optimizer:

  1. Iterates over a matrix of configuration settings (e.g., useHttpCloak: [true, false], injectStealthJS: [true, false], differing Chrome versions).
  2. Executes npx tsx dump-fp.ts zendriver or npx tsx dump-fp.ts stealth --allow-trackers for each permutation.
  3. Parses the output in fp-dump.json, prioritizing:
    • Lowest suspect_score
    • Absence of bot_type: browser_automation_studio or any tampering flags.
  4. Updates app-config.json with the optimal configuration values.

Triggering

Run /goal Optimize fingerprint configuration to achieve a suspect score of 0 on Zendriver. Or Hermes can autonomously decide to run this when success rates drop.

Constraints

  • Golden Template Priority: Stealth (Camoufox) is the primary baseline. Do not degrade Camoufox's strict tracking block (block_webrtc: true) in production just to satisfy a demo script.
  • A/B Testing: Run multiple configurations iteratively. Do not execute parallel dump-fp.ts processes to avoid IP rate-limiting.