Start Here

This is the operator-focused path for understanding and launching StyloBot.

What StyloBot does

StyloBot analyzes each request across multiple detectors, combines those detector contributions into:

• Bot probability (0.0 to 1.0)
• Confidence (0.0 to 1.0)
• Risk band (VeryLow, Low, Elevated, Medium, High, VeryHigh)
• Recommended action (Allow, Throttle, Challenge, Block)

You keep policy control. StyloBot gives evidence, not black-box decisions.

Why teams choose StyloBot

• Speed with intelligence: low-latency request decisions with explainable evidence.
• Signature-first identity: a visitor is represented as a zero-PII signature, not raw IP, UA, or location.
• Time-aware detection: behavior is evaluated across request history, not just single snapshots.
• Higher resolution outcomes: multi-vector fuzzy matching across time improves confidence on evasive traffic.
• Operator control: policy remains yours, including allow/challenge/throttle/block strategy.

5-minute quick path

1. Open /Home/LiveDemo
2. Open /_stylobot in a second tab
3. Call /bot-detection/check with a normal browser User-Agent and then a scraper-like User-Agent
4. Compare probability, risk, top reasons, and suggested action
5. Review detector contribution trends in the dashboard

How request evaluation works

flowchart LR
  A[Request] --> B[Fast-path detectors]
  B --> C[Derived / correlation detectors]
  C --> D[Aggregation + risk band]
  D --> E[Policy mapping]
  E --> F[Allow/Throttle/Challenge/Block]
  D --> G[Dashboard + telemetry]

For a detailed walkthrough of this pipeline, read How StyloBot Works.

Read next

• Running Locally: local setup and smoke tests
• How StyloBot Works: scoring, risk bands, policy flow
• Detectors In Depth: what each detector checks and how to tune it
• Live Demo Guide: how to present and interpret live output
• Deploy on Server: practical production rollout
• Connected Signature Exchange Spec: endpoint contract, delta sync, merge reputation, and zero-PII controls
• GitHub Docs Map: deeper technical docs in the GitHub repo

Safe rollout defaults

• Start in observe mode before taking blocking actions
• Tune on real traffic for at least one full traffic cycle
• Use path/policy-specific thresholds instead of one global aggressive threshold
• Keep secrets managed outside source control