Question 1

How accurate are Scandar's scans?

Accepted Answer

The offline pattern scan has near-zero false positives. The AI behavioral analysis may occasionally flag unusual but benign patterns. Every finding includes matched content and an explanation.

Question 2

Can Scandar miss threats?

Accepted Answer

No scanner catches everything. Scandar detects 20+ threat categories with hundreds of patterns, but novel attacks may evade detection. We continuously update detection rules.

Question 3

What do trust score numbers mean?

Accepted Answer

Trust scores range from 0 to 100. Safe (90-100), Caution (70-89), Risky (40-69), Dangerous (0-39).

Question 4

What's included in the free plan?

Accepted Answer

10 scans per month with full offline + AI behavioral analysis. No account required for your first scan. CLI with offline Layer 1. Free forever.

Question 5

What does Pro add?

Accepted Answer

Unlimited scans, AI Fix auto-remediation, scan history, API access, CLI with AI behavioral analysis, Agent Security Reports, and priority support. $49/mo.

Question 6

What is scandar-guard?

Accepted Answer

A lightweight SDK that wraps your AI client and inspects every message, tool call, and agent response at runtime inside your application process.

Question 7

Does Guard send my prompts to Scandar?

Accepted Answer

No. Guard runs 100% in-process on your infrastructure. Prompts, responses, and tool call contents never leave your environment.

Question 8

How do I install scandar-guard?

Accepted Answer

pip install scandar-guard (Python) or npm install scandar-guard (TypeScript). Wrap your client with one line: client = guard(Anthropic()).

Question 9

Do you store my scanned files?

Accepted Answer

For authenticated users, scan results are stored for history. Original file content is stored temporarily for AI Fix. Unauthenticated scan data is not stored.

Question 10

Is my data used to train AI models?

Accepted Answer

No. We do not use your scan data to train models. Anthropic's API usage policy does not use API inputs for training.

Frequently Asked Questions

Scanning

Trust Scores

Plans & Billing

Runtime Security (scandar-guard)

Privacy & Security