Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 13 additions & 0 deletions docs/METADATA-SCOPE-REPAIR-REVIEW.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
# Independent review: two metadata-only scope corrections

Reviewed2026-10-24 by ci_repair_resume. Approve the exact transit/Shannon metadata transaction for separately gated public review/release. No equation, default, solver, model ID, count, diagram/control, output column, admission row or application engine change is approved.

Independent comparison against committed pre-adoption source found exactly name/description/source/reference changes in each of two files. All998other public files remain byte-identical; cumulative84admission ledger and all22core modules are unchanged. Transit beforeba7fb744f6c0ee69a2653fce8399dcd125ef8af7b3bea795717c469ccbc8b813 becomes e9817e6ac36f8257472d2d70464e8dbfeb6e2c1671e9a0e26509bcd3fdf97c29. Shannon beforebf2c9cb984f264d2e082935523d18f83dc594cefd29c3e98a2eb902f135d4618 becomes394ae80f268be47d553847eb517a80f9896331bbcc0946bb508f9d29ba394a07. Approved current copies exactly match their reviewed fixtures/proposals.

Transit now discloses an authored, uncalibrated feedback illustration rather than claiming the background Taylor paper reproduces its equations. The warning correctly requires positive service, nonnegative ridership/Car Users and ridership<1000*service+1 throughout all evaluated stages; equality divides by zero and exceeding it reverses named transfers. Signed revenue and negative reinvestment remain visible limitations. The corrected author-page DOI is background, not an equation source; inaccessible publisher text is not treated as verified. Shannon now describes its actual capped first-order fluid compartment/outcome budget, illustrative loss fraction and unverified equation provenance, rather than channel capacity/coding, M/M/1 or physical calibration. No denominator clamp, flow repair, parameter rescaling or numerical improvement is implied. IDs/legacy user paths remain.

Independently ran original isolated proposal/source/domain checks:30passed0.61s. Fresh adoption/history/candidate/public-index checks:96passed23.86s, including exact full before/after trajectories for121transit and201message samples. Fresh historical v17/follow-up chain checks:74passed3.85s. Historical helper pins exact two path-before-after tuples, verifies current/source bytes, exact four-field deltas and original archive hashes before returning older bytes. The final follow-up resolver verifies metadata history before its own historical after hash; existing all29transaction regression covers the transit chain. An earlier read of the in-progress helper prompted this check; rereading the final frozen file and fresh tests confirmed the fix was already present. An additional explicit transit chain regression was inspected after this fresh run and adds the same archive-before/v17 assertions. No historical receipt/assertion was rewritten or weakened.

Verified every row of the fresh1000default audit against current files, all22engine bindings and minimal public projection. Author-run audit1000passed92.21s SHA2565d9704abdd552f289614999ce2bac80468a66b0e45fe61f10b31c91eabd816e4; this reviewer verified bindings rather than redundantly rerunning the audit. Public index SHA2562f025b8c08966cf374128d7e14d20eda8185a6a9d0d773c077276ebe37639eef differs from the prior index in exactly two model-hash fields, with unchanged order/counts/statuses/engine metadata and preserved formatting. Final17path author allowlist SHA25616a33fc2c986b3d331c9955f1c0cc243e3961c58402acec1b742b6b610e87063 matches all frozen files.

Public packaging excludes the complete private operator audit and adds this review; fixtures/transaction/docs/tests contain no user data or credentials. Full final public regression/CI and scoped production review remain separate gates. If incomplete at12:00UTC, leave the clean PR unmerged rather than bypass a gate. Finite unchanged trajectories do not certify scientific validity or safe arbitrary parameter exploration.
39 changes: 39 additions & 0 deletions docs/METADATA-SCOPE-REPAIRS.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,39 @@
# Honest scope for two retained teaching models

The1000-model catalog retains these same IDs, equations, controls, settings,
layouts and output columns. Only name, description, source and reference change.
Complete before/after default trajectories are identical: transit121samples,
message processing201samples. All1000defaults passed a fresh full audit; numerical
execution is not scientific certification. No count, solver or equation changed.

public-transit-ridership-service-tra is now an authored transit ridership/service
feedback illustration with a restricted domain. For attractiveness
A=S/100*(1-R/(1000S+1)), loss=.05R/A is undefined at positive R=50000,S=49.999.
Nonnegative transfer meanings require positive S, nonnegative R and Car Users,
and R<1000S+1 throughout every evaluated state, including intermediate integration
stages. The existing equations do not enforce this domain. It is unsuitable for
unrestricted parameter exploration or calibrated policy claims. Signed revenue
is a cumulative balance, not automatically an invalid state. No arbitrary clamp
or replacement law is supplied.

The unrelated direct DOI is removed. Taylor et al. remains background only via
[the UCLA author page](https://taylor.its.ucla.edu/pubs/nature-and-or-nurture-analyzing-the-determinants-of-transit-ridership-across-us-urbanized-areastransportation-research-part-a-policy-and-practice-431-60-77/)
and DOI10.1016/j.tra.2008.06.007; it is not an equation source or calibration claim.
The publisher full text was inaccessible during the recorded source review.

shannon-communication-channel-tec is now Capped-service message processing with
illustrative losses. The fixed60service cap and p=.5exp(-SNR/2) are authored fluid
processing assumptions, not Shannon channel capacity/coding or a calibrated
modulation error law. Below the cap, processing is first-order with implicit
time scale1; this is not an M/M/1 simulation. The useful queue/success/error
budget and original controls are preserved. Source is empty rather than a
fabricated equation citation; the historic Shannon link remains in archived
review evidence and does not establish these equations.

Exact approved after hashes:
transit e9817e6ac36f8257472d2d70464e8dbfeb6e2c1671e9a0e26509bcd3fdf97c29;
message processing394ae80f268be47d553847eb517a80f9896331bbcc0946bb508f9d29ba394a07.
Hash-bound before/after fixtures and metadata transaction preserve earlier
admission/source-review tests without rewriting their historical receipts.
Other legacy source/meaning findings remain explicit backlog; these corrections
do not certify the remaining catalog scientifically or alter saved/private copies.
10 changes: 5 additions & 5 deletions models/technology/shannon-communication-channel-tec.json
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
{
"id": "shannon-communication-channel-tec",
"name": "Shannon Communication Channel Model",
"description": "Claude Shannon's information theory model showing channel capacity, noise, and error rates. Foundation of modern information theory. Reference: Shannon, C.E. (1948). A Mathematical Theory of Communication.",
"name": "Capped-service message processing with illustrative losses",
"description": "Illustrative continuous-stock message-processing compartment with capped service and two cumulative outcome counters. Queue Q receives 50 messages per arbitrary time unit; service is min(Q,60), implicitly using a one-time-unit first-order processing scale below the cap, not an instantaneous-work-conserving queue. Outcomes use a fixed illustrative fraction 0.5*exp(-(10/2)/2); the Signal_Power, Noise_Power and SNR labels are legacy knobs for that chosen fraction, not calibrated physical channel measurements. This is not an M/M/1 queue, Shannon capacity calculation, bit-error model, coding theorem or stochastic packet simulation. Fractional messages are fluid quantities. The default Euler step is 0.5; mathematical balances are tested, not empirical performance. Formerly Shannon Communication Channel Model; the Shannon1948 citation did not establish these service/loss equations. Original equation provenance is unverified. Existing IDs, equations, inputs, diagram and outputs are retained for compatibility.",
"stocks": [
{
"id": "s1",
Expand Down Expand Up @@ -151,8 +151,8 @@
"end": 100,
"dt": 0.5
},
"source": "https://doi.org/10.1002/j.1538-7305.1948.tb01338.x",
"reference": "Shannon, C.E. (1948). A Mathematical Theory of Communication. Bell System Technical Journal, 27(3), 379-423.",
"source": "",
"reference": "Authored illustrative processing-compartment scenario; no original executable source, calibrated service law or physical error-probability law has been established.",
"created_at": "2026-01-15T19:52:37.285857+00:00",
"updated_at": "2026-01-15T19:52:37.285857+00:00"
}
}
10 changes: 5 additions & 5 deletions models/transportation/public-transit-ridership-service-tra.json
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
{
"id": "public-transit-ridership-service-tra",
"name": "Public Transit Ridership and Service Quality Model",
"description": "Model of transit ridership dynamics with service quality feedback and mode choice. Based on transit planning research.",
"name": "Transit Ridership and Service Feedback Illustration",
"description": "Authored, uncalibrated ridership/service feedback illustration. Equations are not verified as a reproduction of Taylor et al. The finite default trajectory is illustrative only, not a forecast or policy assessment. Restricted domain: positive service, nonnegative Car Users, and 0 <= ridership < 1000*service+1 throughout every evaluated state; equality divides by zero and exceeding the threshold reverses the named transfer flows. Initial conditions alone do not ensure this domain remains valid. Parameter exploration can leave it; the runtime does not enforce this restriction. Revenue is a signed cumulative balance; negative balances also reverse the linear service-investment term, without a separate financing model.",
"stocks": [
{
"id": "s1",
Expand Down Expand Up @@ -259,6 +259,6 @@
},
"created_at": "2026-01-15T14:24:00.393635",
"updated_at": "2026-01-15T14:24:00.393644",
"source": "https://doi.org/10.1016/j.tra.2008.01.020",
"reference": "Taylor BD, et al. Nature and/or nurture? Analyzing the determinants of transit ridership across US urbanized areas. Transportation Research Part A. 2009;43(1):60-77."
}
"source": "",
"reference": "Authored illustrative equations; parameter and equation provenance unresolved. Background only: Taylor BD, Miller D, Iseki H, Fink C (2009), Nature and/or nurture? Transportation Research Part A 43(1):60-77. https://doi.org/10.1016/j.tra.2008.06.007. Not an executable equation source."
}
46 changes: 46 additions & 0 deletions operations/library-metadata-scope-20261024.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,46 @@
{
"scope": "Independently approved metadata-only honest source/domain scopes; no math/count changes",
"active_count": 1000,
"replaced_count": 2,
"retired_count": 0,
"admitted_count": 0,
"entries": [
{
"file": "models/transportation/public-transit-ridership-service-tra.json",
"id": "public-transit-ridership-service-tra",
"before_sha256": "ba7fb744f6c0ee69a2653fce8399dcd125ef8af7b3bea795717c469ccbc8b813",
"after_sha256": "e9817e6ac36f8257472d2d70464e8dbfeb6e2c1671e9a0e26509bcd3fdf97c29",
"archive": "tests/fixtures/metadata-scope-20261024/public-transit-ridership-service-tra-before.json",
"source": "tests/fixtures/metadata-scope-20261024/public-transit-ridership-service-tra-after.json",
"changed_fields": [
"name",
"description",
"source",
"reference"
],
"points": 121,
"full_trajectory_unchanged": true,
"output_sha256": "bc5d134cc6571aa1dc5ad39a57bd03382736a9a7cb11256d8d52499b0f8da55f"
},
{
"file": "models/technology/shannon-communication-channel-tec.json",
"id": "shannon-communication-channel-tec",
"before_sha256": "bf2c9cb984f264d2e082935523d18f83dc594cefd29c3e98a2eb902f135d4618",
"after_sha256": "394ae80f268be47d553847eb517a80f9896331bbcc0946bb508f9d29ba394a07",
"archive": "tests/fixtures/metadata-scope-20261024/shannon-communication-channel-tec-before.json",
"source": "tests/fixtures/metadata-scope-20261024/shannon-communication-channel-tec-after.json",
"changed_fields": [
"name",
"description",
"source",
"reference"
],
"points": 201,
"full_trajectory_unchanged": true,
"output_sha256": "7d14de9498fae2885595250175dcc8835f23009445426185a4163fa83fd6211d"
}
],
"baseline_public_receipt": "tests/fixtures/metadata-scope-20261024/library-validation-before.json",
"baseline_public_receipt_sha256": "200cca04903c9ad03c051c89d04d58aa4ef05f49d9a80dfa5feab285dea4825e",
"admission_ledger_sha256": "639d24fcb4a00e2d8f3b205474a0830b3bc8dacb3754e314adebc35e1710d1c6"
}
4 changes: 2 additions & 2 deletions static/library-validation.json
Original file line number Diff line number Diff line change
Expand Up @@ -4927,7 +4927,7 @@
},
{
"id": "shannon-communication-channel-tec",
"sha256": "bf2c9cb984f264d2e082935523d18f83dc594cefd29c3e98a2eb902f135d4618",
"sha256": "394ae80f268be47d553847eb517a80f9896331bbcc0946bb508f9d29ba394a07",
"status": "passed"
},
{
Expand Down Expand Up @@ -5002,7 +5002,7 @@
},
{
"id": "public-transit-ridership-service-tra",
"sha256": "ba7fb744f6c0ee69a2653fce8399dcd125ef8af7b3bea795717c469ccbc8b813",
"sha256": "e9817e6ac36f8257472d2d70464e8dbfeb6e2c1671e9a0e26509bcd3fdf97c29",
"status": "passed"
},
{
Expand Down
Loading
Loading