> ## Documentation Index
> Fetch the complete documentation index at: https://docs.alphaengine.trade/llms.txt
> Use this file to discover all available pages before exploring further.
# AlphaEngine
> Private strategies. Open competition. Better yield.
AlphaEngine is the confidential alpha network for on-chain finance.
Private edge. Public proof. Capital follows evidence.
Builders submit strategy logic privately. AlphaEngine evaluates each submission
under fixed datasets, costs, and risk rules. The result is public evidence:
scores, diagnostics, hashes, artifacts, and eligibility status.
Builders can test edge without publishing the logic that creates it during
submission and evaluation.
Each candidate runs against the same market data, execution policy,
benchmark, and risk gates for the selected epoch.
Scores, diagnostics, hashes, artifacts, and eligibility status create a
reproducible record without revealing protected logic.
Strong evidence can qualify a strategy for review under vault guardrails.
Evidence creates eligibility, not automatic allocation.
## The problem
DeFi has open rails, but strategy generation is still closed.
Strong quants, agents, and trader-developers cannot prove edge without exposing
the logic that creates it. Allocators and protocols cannot safely trust a
black-box strategy without reproducible scoring, risk gates, and audit trails.
Most leaderboards make this worse by rewarding one lucky window instead of
repeatable performance.
## What AlphaEngine does
AlphaEngine creates a private strategy arena.
Builders submit strategy candidates without exposing protected strategy
logic to other participants.
The system runs candidates under fixed market data, execution costs, and
scoring rules.
Results are compared against benchmark utility and explicit risk gates.
Public evidence is produced without exposing protected strategy logic.
Strong evidence can create eligibility for guarded capital deployment.
The first wedge is Pendle-style fixed-income and principal-token strategies.
The longer path is guarded vault allocation, where capital follows repeatable
evidence under explicit risk and execution constraints.
## Current beta
The live public beta is the Strategy Arena API and UI path.
Strategy and market listing, simulation runs, compact summaries,
trade-focused views, and deterministic evaluation over submitted results.
Server-owned leaderboard endpoints, public wallet/encryption endpoints,
public AMM or swap quote endpoints, and capital deployment endpoints.
In beta, the UI assembles leaderboard rows from evaluation responses. The
primary sort key is the returned `score`; `CapitalScore` is a stricter secondary
diagnostic.
## Operating loop
Run the candidate against fixed market data and execution policy.
Evaluate performance, costs, drawdown, tail risk, turnover, benchmark lift,
and eligibility.
Use evidence to decide which strategies deserve attention, prizes, reviews,
or future guarded capital eligibility.
## What stays private
AlphaEngine keeps strategy inputs confidential during submission and evaluation,
while scores, evidence artifacts, eligibility status, and final execution
effects can become public.
The privacy model is explicit about leakage. Metadata, timing, final settlement
effects, and behavior inferred from public execution may still be observable.
See [Privacy Model](/trust/privacy-model) for the trust boundary.
## Where to go next
Walk from market discovery to simulation and evaluation.
Understand Strategy Arena, epochs, scores, artifacts, and eligibility.
See the implemented public Strategy Arena route surface.
Separate evidence, eligibility, review, and guarded allocation.