Privacy scoring
Run a privacy check at any time through the agent’s chat interface.How the score is calculated
The check starts at 100 points and deducts for observed risk patterns. A score of 90 or above is healthy. Below 70 indicates significant exposure.| Condition | Deduction | Severity |
|---|---|---|
| More than 5 unspent stealth addresses | −10 | Medium |
| All payment amounts identical | −15 | High |
| Consecutive payments less than 60 seconds apart | −20 | High |
| Never withdrawn any payments | −5 | Info |
| Connected wallet is the agent address | −5 | Info |
What to avoid
1. Withdrawing to the same address
Sending multiple stealth address withdrawals to a single known wallet lets any observer link all of those payments to you. Bad:2. Timing correlation
Withdrawing from multiple stealth addresses within seconds of each other creates a timing cluster that observers can use to group the addresses together. Bad:3. Amount fingerprinting
Sending the exact same amount repeatedly creates a pattern that makes your payments distinguishable even across different stealth addresses. Bad:4. Address reuse
Stealth addresses are one-time by design — the Wraith protocol generates a fresh address for every payment automatically. If you are building custom integrations using the EVM crypto primitives directly, callgenerateStealthAddress() for each payment rather than reusing a previous output.
5. Linking on-chain identity
Withdrawing from a stealth address directly to a wallet tied to your ENS name,.wraith name, or any other on-chain identity defeats stealth entirely.
Bad: Withdraw stealth funds directly to a wallet that has ever interacted with your identity.
Good: Use an intermediate address with no on-chain identity, then move funds from there.
How the agent helps
The Wraith AI agent is privacy-paranoid by design. It acts before you make a mistake, not after.Warns before risky actions
Explains the specific risk before executing any operation that could degrade privacy.
Suggests alternatives
Recommends fresh addresses, time spacing, and amount variation alongside every warning.
Runs proactive checks
Analyzes your full activity history and flags emerging patterns before they become exposures.
Respects your decision
Executes the operation after warning if you confirm — it advises, it does not block.
Example: agent warning on bulk withdrawal
Example: proactive amount fingerprint detection
Best practices summary
| Practice | Why it matters |
|---|---|
| Use a fresh destination for each withdrawal | Prevents linking stealth addresses to one identity |
| Space withdrawals at least 1 hour apart | Defeats timing correlation analysis |
| Never withdraw to your connected wallet | Keeps your identity separate from stealth activity |
| Vary payment amounts slightly | Prevents amount-based fingerprinting |
| Use different times of day | Avoids timezone-based profiling |
| Consolidate stealth addresses periodically | Reduces on-chain footprint and keeps score high |
Your privacy score is a snapshot of current activity patterns, not a guarantee of anonymity. Follow these practices consistently — a single careless withdrawal can link otherwise-private payments.
