Assets tracked
250+
Public market universe with factorized score context.
Loading page
Bringing in the next surface without the heavy transition shell.
Prism exposes provider status, asset coverage, cache keys, stale windows, model writes, and fallback behavior before users trust a score or AI brief.
Assets tracked
250+
Public market universe with factorized score context.
Categories
14
Market segments visible in the coverage map.
Chain lanes
4
Token and wallet lanes prepared for deeper indexing.
market
Live 60s
Top market rows refresh from the public provider and fall back to the local Prism model when needed.
token
Warm 8m
Scores are factorized and can survive provider outages with caveats.
wallet
Provider pending
Wallet movement is modeled while chain-indexer credentials and routes are attached.
alert
Delayed
Compound alerts degrade into review mode if evaluation misses cadence.
research
Fresh
Saved briefs and exports write immediately.
Ranks tokens, candles, volume, categories, and liquidity.
Freshness: < 2 minutes
Fallback: Use local Prism model when the public market provider is unavailable.
Coverage: Top 250 public crypto assets across majors, L1, L2, DeFi, AI, meme, stable, infrastructure, gaming, RWA, payments, privacy, exchange, and social categories.
Normalizes transfers, counterparties, wallet activity, and graph edges.
Freshness: < 5 minutes
Fallback: Hold latest known wallet snapshot and mark copy-risk stale.
Coverage: Ethereum, Solana, Base, Bitcoin, and multi-chain wallet routes prepared for provider expansion.
Scores liquidity quality, holder pressure, route quality, and anomaly pressure.
Freshness: < 10 minutes
Fallback: Keep previous score, expose affected factor, and disable confident AI language.
Coverage: All visible Prism token rows receive a factorized score.
Stores cited briefs, source chips, caveats, room notes, and export records.
Freshness: Immediate writes
Fallback: Draft-only room with local export disabled until persistence returns.
Coverage: Team rooms, saved memos, and PDF report exports.
Evaluates compound rules against market, wallet, flow, and risk snapshots.
Freshness: < 1 minute for watched rules
Fallback: Batch evaluate on next request and label delayed rules.
Coverage: Compound rules model source, cadence, and severity.
Methodology rule: AI briefs can cite only visible panels. If any layer is stale, Prism must lower confidence, name the affected layer, and route the user back to source evidence.