​
Four profiles

What you get: population-corrected trader rankings (Deflated Sharpe, not raw), temporal alpha decomposition that answers where in time a trader’s edge lives, and a non-custodial execution primitive built on scoped EIP-712 signatures with on-chain enforced caps.Why it matters: a headline Sharpe cannot distinguish a trader with genuine entry timing from one who held through a favourable regime. The decomposition separates them and attaches a permutation p-value to every claim, so a 40-trade account is not labelled a “funding arbitrageur” by accident.Start: Score a trader → Temporal alpha
What you get: a provider reliability leaderboard that scores your existing vendors on whether their answers were right, not whether their API was up. Plus wallet clustering with per-edge provenance.Why it matters: most vendor evaluations measure uptime. This measures accuracy against ground truth, and publishes the exclusions — only three of six named providers are structurally scoreable, and the board says so rather than quietly ranking six.Start: Provider reliability
What you get: Sentry authority-change detection for monitored contracts, evidence bundles that record which sources produced nothing and why, and an audit log across workspace actions.Why it matters: an evidence bundle that distinguishes “checked and clean” from “never checked” is the difference between a defensible file and an indefensible one.Start: Sentry alerts → Security overview
What you get: 38 read-only MCP tools covering wallets, signals, memecoins, providers, and on-chain data, usable from Claude Code, Claude Desktop, Cursor, or Windsurf.Why it matters: the tool layer is read-only by construction, not by promise — every capability resolves to a reading function and clustering is called with persist=False, asserted by a test. An inferred edge written into the entity graph on a language model’s say-so would be a write reachable from a prompt.Start: MCP server → Claude setup

​
Where the honest answer is “use something else”

This is a narrow instrument. It is not a Bloomberg replacement and does not attempt the coverage of a mature vendor.

​
The honest baseline

Roughly 51% of copy-traders lose money, and only about 44% of leaders make money for the people copying them. Scienter is built to maximise the odds of the narrow honest case, not to promise an edge that does not exist. Every ranking is designed to make the absence of skill as visible as its presence.
Scienter publishes impersonal market signals of general and regular circulation. Nothing here is investment, legal, or tax advice, or a recommendation to buy, sell, or hold any asset. Scienter is not a registered investment adviser or broker-dealer. Trading digital assets can result in the total loss of your funds — see Disclaimers for the full text.