QuantLite firewall: provenance-aware, issuer-independent scoring - #1
Merged
Conversation
Introduces quantlite.score, an open scoring methodology (QLS-1.0) for trading track records, built on the existing forensics and resampling stack: - Composite 0-100 score: Deflated Sharpe skill (35%), bootstrap robustness (20%), tail risk (20%), consistency (15%), track record sufficiency (10%), with letter grades A+ through F. - Integrity checks that flag return smoothing, outlier dependence, the short-volatility signature, cherry-picked start dates, and short records. Critical flags cap the score at 40, warnings at 70. - Verifiable artifacts: canonical JSON with SHA-256 input digest and content hash; verify_artifact() reproduces any score bit for bit. - Deterministic by construction: seeded bootstrap, platform-independent canonicalisation, calibration constants frozen per methodology version. - Full specification in docs/score.md, 23 new tests, changelog and README updates.
The score answers what the returns were; the firewall answers where they came from. A SourceAttestation binds an input digest to an independent origin (exchange custody, fund admin, prime broker, allocator) and the party vouching for it. A score is firewall-clean only when its inputs are attested to a source the rated manager does not control, making the rating independent of who pays. - score.provenance: DataSource/is_independent, SourceAttestation, AttestedScore/FirewallReport, assert_firewall/FirewallError. - Five-check verification in one pass: artifact consistency, attestation consistency, digest agreement, score reproduction, source independence. - 24 new tests; docs/firewall.md spec with marketplace-operator integration; STRATEGY.md commercial rationale; changelog.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The firewall: provenance-aware, issuer-independent scoring
The QuantLite Score already answers what a track record's returns were (a
reproducible artifact with a SHA-256 input digest). This adds the second,
load-bearing question: where did the returns come from?
A score is only worth money while the rated party cannot buy a better one.
This layer makes the score independent of who pays — enforced in code, not
policy.
What's in it
score.provenance—DataSource/is_independent,SourceAttestation,AttestedScore/FirewallReport,assert_firewall/FirewallError.SourceAttestationbinds an input digest to an independent origin(exchange custody, fund administrator, prime broker, allocator) and the
party vouching for it, sealed with its own tamper-evident hash.
consistency, digest agreement, score reproduction, source independence.
firewall-clean. The report distinguishes "the number is fake" from "the
number is real but the provenance is not independent."
Quality
engine hashing untouched.
Docs
docs/firewall.md— specification plus the marketplace-operatorintegration (venue as custodian of record).
STRATEGY.md— commercial rationale and the non-negotiable invariant:paying more can never produce a better number.