RugSignal, clearly explained.
RugSignal is a Solana token-risk scanner for traders, researchers and developers. It presents timestamped contract, wallet and liquidity evidence with explicit uncertainty. It does not certify tokens as safe or execute trades.
Public capability guide · Updated 2026-10-18
This is a quick capability map. The methodology explains individual rules; the guides show how to investigate them. The accuracy record separates observations from reviewed outcomes.
What is RugSignal?
RugSignal is a Solana token-risk scanner for traders, researchers and developers. It presents timestamped contract, wallet and liquidity evidence with explicit uncertainty. It does not certify tokens as safe or execute trades.
Which blockchain and token types does it support?
Solana only. Supported parsing covers legacy SPL mints and known Token-2022 extensions. Unsupported extensions or pool mechanics remain unknown. Canonical Wrapped SOL and USDC use identity-specific reference context; a matching ticker alone does not qualify. No other chain or project access token is launched.
What risks can it assess?
Available evidence includes mint/freeze authorities, transfer-fee and supported Token-2022 controls, holder concentration, deployer exposure and attributable prior launches, sampled funding relationships, early recipients and supported liquidity control. A capability or association is not proof of malicious intent; unavailable history stays unknown.
How does the deterministic evidence engine work?
A versioned rules engine turns supported observations into explicit warning signals. Reports distinguish raw references, parsed facts, inferences and assessment. Blockers cannot be cancelled by positive signals. Read the source, timestamp, parser/scoring version and sampling limitations for a claim. AI explanations may be unavailable; computed evidence remains available independently.
How are wallet relationships interpreted?
Observed transfers and shared funders are investigative leads, not proof that wallets share an operator. Exchange withdrawals, services, pools and incomplete samples can create ambiguous links. Concentration may be a sampled lower bound; token accounts are not people. A transfer is not a confirmed sale without supported swap evidence.
What do risk levels and scoring applicability mean?
Low, Elevated, High and Severe are deterministic policy categories, not calibrated fraud probabilities. Incomplete evidence cannot produce Low. Canonical native/issuer reference assets have a separate model-applicability context; ordinary-token liquidity ratios or funding heuristics are not applied indiscriminately. A low numeric score can coexist with incomplete evidence and an Elevated policy floor.
What does incomplete evidence or stale data mean?
Evidence coverage describes available checks, not detection accuracy. Unknown checks are not passes. Each saved observation has its own timestamp; it is not a live feed. Immutable share reports never refresh themselves. Latest-token pages are indexable only under the existing freshness and evidence policy. Recheck deliberately when new evidence is needed.
Can RugSignal detect every rug pull?
No. Future intent, abandonment, new mechanisms, unsupported pools and unavailable history can escape on-chain checks. A lock, revoked authority, quote or risk category is not a safety or exit guarantee. Independent predictive accuracy remains unverified: unresolved observations and synthetic software tests do not establish precision or recall.
What is available for free?
Basic scans accept a Solana mint address or supported token link without connecting a wallet. Saved reports and share cards show supported deterministic signals and limitations. Visiting a saved report does not start a scan, quote or AI request. Educational guides and methodology are public.
What additional Pro and monitoring functions exist?
Entitled accounts can use deeper investigations, sampled cluster views, watchlists and risk-change monitoring with linked Telegram alerts. Wallet authentication signs a message, not a spending approval. Published website terms describe 30-day access at 0.2 SOL or 19 USDC; public checkout is currently closed pending real payment acceptance. Holder access also depends on configured token settings; there is no launched RugSignal token. Server-side entitlements and configured limits apply.
What developer integrations exist?
The public developer overview describes the versioned RiskAPI, TypeScript SDK and signed change-event contracts. Live B2B access requires commercial rights and account approval. Integration interest does not create a key or activate access. Public documentation is separate from authenticated APIs; simulation is not a real partner integration.
How is detection accuracy evaluated?
Factual correctness and pre-event warning performance are separate questions. Evaluation requires independently evidenced event times, frozen pre-event facts, survivors and ordinary failures, held-out cohorts and related-deployer separation. Precision, recall, false-positive rate, misses, lead time, coverage and uncertainty require reviewed labels. No independently reviewed predictive benchmark or superiority claim is currently published.
What liquidity and trading limitations matter?
Supported Raydium V4 LP and Pump curve evidence is protocol-specific. Other withdrawal mechanics may remain unknown. Migration is not inherently a rug. Bounded exit-size quotes are informational, expire quickly and do not execute trades or guarantee fills. Wallet holdings, market capitalization and displayed liquidity do not establish the price at which an entire position can exit. Private sessions, watchlists and billing are not public reports.