Asset A
Deep marketQuoted price10
5% executable depth500,000 XLM
Liquidity risk intelligence for Stellar
A price is only as credible as the liquidity behind it.
Keel measures executable market depth, estimates how costly it is to move a market, and surfaces collateral risk before thin liquidity is treated like deep liquidity.
Read-onlyOpen methodologyNo wallet connection
Market depth profileIllustrative view
The missing measurement
Price is not liquidity.
A market can print a price with very little executable depth behind it. For lending and collateral decisions, the question is not only what the price is, but how much can actually trade near it.
Asset B
Thin marketQuoted price10
5% executable depth800 XLM
Same quoted price. Very different collateral risk.
The evidence layer
What Keel measures
The output is designed for inspection. Each measure keeps its unit, source, and uncertainty close to the conclusion.
Effective liquidity depth
Measure quote-asset value that can trade before the market moves by ±2%, ±5%, and ±10%.
BUY + SELL · QUOTE ASSETManipulation resistance
Estimate the notional needed to reach defined price targets, together with whether each target is reachable.
COST + REACHABILITYMaximum safe collateral
Surface a conservative collateral recommendation derived from executable depth and manipulation constraints.
CONSERVATIVE RECOMMENDATIONSupporting risk signals
Keep spread, source divergence, genuine activity, and holder concentration in view where data is available.
PROVENANCE AWAREFrom market to finding
A clear path from observation to evidence.
Keel keeps the chain visible without asking visitors to decode blockchain internals first.
Observe the market
Keel reads Stellar market data from SDEX, AMMs, and supporting observations.
Apply the methodology
Depth, manipulation, collateral, and rule-based flags turn observations into findings.
Publish evidence
Results stay inspectable through the dashboard, API material, and historical case studies.
Risk, with its working shown
No mystery score. Every flag can be inspected.
Keel uses rule-based bands rather than an unexplained weighted score. The finding comes first, then the triggered checks, the checks that could not be evaluated, and the provenance behind the result.
Partial data never means clear data.
Asset findingCritical
TriggeredZero executable depth within 2%Manipulation target is reachable at low cost
Not evaluatedHolder concentration
Partial data never means clear data. Every result keeps the checks that could not be evaluated visible.
Historical evidence
A known incident, replayed as evidence.
The February 2026 USTRY/Blend incident is Keel's primary historical test case. The case study asks whether market-depth and manipulation metrics would have exposed structural risk before a quoted price was treated as large collateral value.
Open the case studyQuoted value looks available.
Price visibility alone does not show how much volume can trade near that quote.
Depth and reachability add context.
Executable depth and manipulation cost make structural weakness legible as evidence.
USTRY / Blend incident.
The historical page documents what the data can support, plus the limits of the reconstruction.
A product built to be checked
The evidence stays close to the claim.
Traceable
Results carry ledger sequence, methodology version, timestamps, and data source.
Reproducible
The methodology and supporting evidence are public and reviewable.
Read the methodologyRead-only
Keel never signs or submits a Stellar transaction. It reports evidence for people to inspect.
View the API contractSee the market behind the price.
Start with the monitored assets, then follow the evidence into methodology and historical context.