diff --git a/PERFORMANCE.md b/PERFORMANCE.md
index 92b4fa8..dae4ff1 100644
--- a/PERFORMANCE.md
+++ b/PERFORMANCE.md
@@ -135,6 +135,14 @@ The fixed paper experiment is stricter than the general report: it accepts no
 missing price rows, performs no forward fill, and requires at least 274 signal
 warm-up sessions before the evaluation window.
 
+`scripts/release_paper_artifacts.sh` is intentionally bound to the reviewed
+input digest. When release-critical paths are unchanged, it regenerates from the
+recorded source commit and timestamp. Changed experiment, report, or paper source
+requires an explicit `QUANTCORTEX_GENERATED_AT` and regenerates from current
+`HEAD`. The wrapper fails rather than labeling a different matrix with the
+reviewed provider provenance. A new panel requires a new, explicitly reviewed
+experiment configuration and source record.
+
 For an explicitly requested live download:
 
 ```bash
diff --git a/docs/evaluation-contracts.md b/docs/evaluation-contracts.md
index 2dddbca..9df5725 100644
--- a/docs/evaluation-contracts.md
+++ b/docs/evaluation-contracts.md
@@ -37,8 +37,14 @@ the complete symbol set and gross limit:
 }
 ```
 
-`target_tape_to_payload` and `target_tape_from_payload` implement that envelope;
-`schemas/canonical_target_tape.schema.json` specifies it for other engines.
+`target_tape_to_payload` and `target_tape_from_payload` implement that envelope.
+`schemas/canonical_target_tape.schema.json` specifies its serialized structure
+and primitive constraints for other engines. Runtime validation additionally
+enforces cross-record portfolio invariants: no duplicate timestamp-symbol rows,
+the declared symbol universe at every decision, and the per-decision gross
+limit. The paper experiment round-trips each variant through this boundary
+before backtesting and publishes canonical payload hashes in
+`paper/results/target_tape_hashes.json`.
 
 The evaluation contract also records post-overlay exposure rules and the
 paper-trading order-state policy: persist intent before submission, block
diff --git a/docs/img/performance_manifest.json b/docs/img/performance_manifest.json
index 753997e..96b3c53 100644
--- a/docs/img/performance_manifest.json
+++ b/docs/img/performance_manifest.json
@@ -1,19 +1,19 @@
 {
   "schema_version": 4,
-  "generated_at": "2026-06-16T22:00:41Z",
+  "generated_at": "2026-06-18T22:06:35Z",
   "generator": {
     "path": "scripts/generate_report.py",
     "script_sha256": "b536aa7fc5e4fe7df6c7ff28c0992629a489869eaec46486db7aff1cb946099b",
     "git": {
-      "source_commit": "d185c44928af865882c4d830066432dc97eb9972",
+      "source_commit": "ef6e7e3414aed0ae9b77d50748a22823529486af",
       "worktree_clean_at_start": true
     },
     "source_tree": {
-      "sha256": "a72c2fcd825bb15a2879a3b33d63487a34348b837ff733b7c1ee9d0232ed19cb",
+      "sha256": "f964d653b5e645e1efe97889e8acb43fbb0c2c8ddeb0835803f07c299abff5a0",
       "file_count": 107,
       "files": {
-        "poetry.lock": "fa8904674c2b24f3141acae607fd0a49d97125bc109c18def9fccbb2256c8de0",
-        "pyproject.toml": "175abd743d38f6b6c1cc8ae42471d62d83c86beb245d93fea231c666f51b2643",
+        "poetry.lock": "abce2cd132651851d81034dea085d7dbfbe5e44321f2e13997db69f46d7f080b",
+        "pyproject.toml": "eaeeb454c28bf7f6d9e530002bb7e88624b56b6c3e1fcb71e6414045cb9c42a0",
         "quantcortex/__init__.py": "14bf1ebdacd054c3738e4704d33da6709a39206463df8b8ced5376da342c4036",
         "quantcortex/alpha/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
         "quantcortex/alpha/factors/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
@@ -37,7 +37,7 @@
         "quantcortex/alpha/validation/alphalens_report.py": "8463289117add78a576e37a5b99a539be5c808fca86100e26deb29e8060aa60d",
         "quantcortex/alpha/validation/factor_decay.py": "9e6e049165f014db2122d9ed57415e45e48142b475dfb7221247d447d9a50397",
         "quantcortex/backtest/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
-        "quantcortex/backtest/conformance.py": "e169cc67cac9eb56d8da4e4a0da1e257bb6cb8d1745baca681f50fab6c62fc57",
+        "quantcortex/backtest/conformance.py": "22f9b18101f0d9adcae98297e395c1ab6a90b18e411f24e7456f1d166b7df9b5",
         "quantcortex/backtest/costs/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
         "quantcortex/backtest/costs/transaction_costs.py": "48200419f37004c58df6f89041e31516d9ad98f66c22f75d11a0b9342474d5c7",
         "quantcortex/backtest/engines/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
@@ -58,7 +58,7 @@
         "quantcortex/backtest/validation/multiple_testing.py": "6b68f6f150dcc8fb90213dcf769307454c0c506e99666a533d5d07ae3ccb0b5d",
         "quantcortex/backtest/validation/survivorship_check.py": "5b772bd6b40ffe44f590ec0ddb00b7e3dc61a1aba9cae8831227325f4f53beff",
         "quantcortex/data/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
-        "quantcortex/data/local_csv.py": "30c11c3200530810b3345f8e29fdf966f35671e30c94fc5bb1b2db656e6df79b",
+        "quantcortex/data/local_csv.py": "6e5f3d24b1477ce122a1e80899e841fffd6d857034189f7d8605705f52c6839c",
         "quantcortex/data/processors/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
         "quantcortex/data/processors/adjustments.py": "ae2d4d0362e3484b89b5a5c57a7fd6320973b208ee0cd5a6be9a3f29d2544e2c",
         "quantcortex/data/processors/calendar.py": "e591bccc3e7f632501b068e7abec1ece182d0ef930589381776fe89158d1df51",
diff --git a/paper/README.md b/paper/README.md
index a982b1e..027c735 100644
--- a/paper/README.md
+++ b/paper/README.md
@@ -38,11 +38,16 @@ scripts/release_paper_artifacts.sh \
 ```
 
 The input must contain `QQQ`, `VGT`, `GLD`, `TLT`, `SPY`, `VIG`, and `SHV` and
-must match the SHA-256 digest in `results/manifest.json` for exact
-reproduction. The wrapper requires committed tracked source, regenerates the
-reviewed `docs/img/` gallery and paper experiment in a detached clean worktree
-with the input mounted outside it, verifies the recorded source commit and clean
-start state, then copies reviewed artifacts back. The
+must match the SHA-256 digest in `results/manifest.json`; the release wrapper
+rejects another matrix rather than attaching the reviewed provenance to it. It
+uses the recorded source commit and timestamp when all release-critical paths
+are unchanged, so the final artifact commit can reproduce without timestamp-only
+drift. Changed experiment, report, or paper source must set
+`QUANTCORTEX_GENERATED_AT` explicitly; that creates a release from current
+`HEAD`. The wrapper requires committed tracked source, regenerates the reviewed
+`docs/img/` gallery and paper experiment in a detached clean worktree with the
+input mounted outside it, verifies the recorded source commit and clean start
+state, then copies reviewed artifacts back. The
 fixed experiment requires complete rows, performs no forward fill, and rejects
 fewer than 274 pre-evaluation sessions. Raw provider data is not committed.
 Aggregate tables, generated LaTeX values, figures, the explicit experiment
@@ -58,6 +63,9 @@ and 63-session sensitivity results. `sharpe_uncertainty.csv` directly resamples
 the conventional sample Sharpe statistic. `comparator_diagnostics.csv` records
 the causal target-exposure comparator after its own costs, while
 `evaluation_contract.json` records the machine-readable semantics.
+`target_tape_hashes.json` records the canonical payload hash, decision count,
+record count, and symbol set for every audited strategy variant without
+publishing the underlying provider matrix.
 
 The primary accounting path is the event-driven engine. It holds explicit
 adjusted-close pseudo-shares between rebalances, sizes targets against post-cost
diff --git a/paper/build_manifest.json b/paper/build_manifest.json
index d766ab7..c1b1f14 100644
--- a/paper/build_manifest.json
+++ b/paper/build_manifest.json
@@ -1,18 +1,18 @@
 {
   "anonymous_pdf": {
     "path": "quantcortex_audit_anonymous.pdf",
-    "sha256": "1ee5d885f03316f70c06dd0c9a7e2cd18ef352f83a7045077925c1f767bf20f9"
+    "sha256": "3bf902d420224032fa11492ddb4f8d3b636972f4a3d7740e4b7aef4bbf8484ff"
   },
   "pdf": {
     "path": "quantcortex_audit_neurips2026.pdf",
-    "sha256": "d43a4c75a1b49b3d2c5e703128e5208be60688e8b1d3bd20c51a3ebbb29e6ede"
+    "sha256": "59fcd30467d449984118d746417d810757d32fbceb72af25d0d4a84a65203000"
   },
   "schema_version": 1,
-  "source_commit": "d185c44928af865882c4d830066432dc97eb9972",
-  "source_date_epoch": 1781647230,
+  "source_commit": "ef6e7e3414aed0ae9b77d50748a22823529486af",
+  "source_date_epoch": 1781820388,
   "source_manifest": {
     "path": "quantcortex_audit_neurips2026.sources.sha256",
-    "sha256": "11096197aee3ed3a6b5505788756789b5deeddee140d3a204c4ed6ac4349b919"
+    "sha256": "ffd76bc58b7b0dd3499cb934a33abd613477478b3fb1fadabb16d5b47b5cbbde"
   },
   "tectonic_bundle": {
     "name": "default_bundle_v33.tar",
diff --git a/paper/main.tex b/paper/main.tex
index 28740c0..c818483 100644
--- a/paper/main.tex
+++ b/paper/main.tex
@@ -319,7 +319,9 @@ \subsection{Reference implementation and order-state controls}
 versions, dependency lock, input SHA-256, experiment design, source-tree hashes,
 and the hash of every published table and figure. A versioned JSON contract and
 canonical long-form target-tape schema make the engine boundary machine
-readable; deterministic fixtures exercise it without using the empirical price
+readable. The fixed experiment round-trips every strategy decision stream
+through the versioned payload boundary and records a canonical payload hash;
+deterministic fixtures test the same boundary without using the empirical price
 matrix. The empirical sections below exercise the
 research and backtest contracts. Order-state controls are verified by
 deterministic tests: paper-trading state is written atomically as a versioned
diff --git a/paper/quantcortex_audit_anonymous.pdf b/paper/quantcortex_audit_anonymous.pdf
index 17f5422..c0d530e 100644
Binary files a/paper/quantcortex_audit_anonymous.pdf and b/paper/quantcortex_audit_anonymous.pdf differ
diff --git a/paper/quantcortex_audit_anonymous.sha256 b/paper/quantcortex_audit_anonymous.sha256
index 00aedb3..2a0a59c 100644
--- a/paper/quantcortex_audit_anonymous.sha256
+++ b/paper/quantcortex_audit_anonymous.sha256
@@ -1 +1 @@
-1ee5d885f03316f70c06dd0c9a7e2cd18ef352f83a7045077925c1f767bf20f9  quantcortex_audit_anonymous.pdf
+3bf902d420224032fa11492ddb4f8d3b636972f4a3d7740e4b7aef4bbf8484ff  quantcortex_audit_anonymous.pdf
diff --git a/paper/quantcortex_audit_neurips2026.pdf b/paper/quantcortex_audit_neurips2026.pdf
index 6514369..0ad1de7 100644
Binary files a/paper/quantcortex_audit_neurips2026.pdf and b/paper/quantcortex_audit_neurips2026.pdf differ
diff --git a/paper/quantcortex_audit_neurips2026.sha256 b/paper/quantcortex_audit_neurips2026.sha256
index 5f31dde..4f8efdd 100644
--- a/paper/quantcortex_audit_neurips2026.sha256
+++ b/paper/quantcortex_audit_neurips2026.sha256
@@ -1 +1 @@
-d43a4c75a1b49b3d2c5e703128e5208be60688e8b1d3bd20c51a3ebbb29e6ede  quantcortex_audit_neurips2026.pdf
+59fcd30467d449984118d746417d810757d32fbceb72af25d0d4a84a65203000  quantcortex_audit_neurips2026.pdf
diff --git a/paper/quantcortex_audit_neurips2026.sources.sha256 b/paper/quantcortex_audit_neurips2026.sources.sha256
index 66a33e0..0752c99 100644
--- a/paper/quantcortex_audit_neurips2026.sources.sha256
+++ b/paper/quantcortex_audit_neurips2026.sources.sha256
@@ -1,9 +1,9 @@
-bbd36070a922fae4569f654b6b905a710c3dc8c318636dfa934ea82842e57d9a  main.tex
+26c3a3aac4ded1b4db38503857a8627d6d7b95d851f8b11897f303d4c6e01618  main.tex
 9ef5395fcb6993271813e21c97ea594bfc45b5f8dcc9ec52fc4e16a809a0091a  anonymous.tex
 4bb1dc333bc0b23fb706bc5755b2cef765a2757284798d6ed9d090d530b47f6e  checklist.tex
 c0c907c664fe2629306b7c07973747b44dc1c04f9ead03ffd292719374e1f531  references.bib
 0c1ad36961fcd9198dcc2558cf2793e1df39973bde8264fd701f5e7970672757  neurips_2026.sty
-f14a548364f13b40e75d44c2eb2c3aa0965668ac35a6acd30d22ef207abb1fdf  results/generated_values.tex
+98611c031c54382859488367dc1f5474a283014e3db2867f8210dab49ac85f25  results/generated_values.tex
 83aea7fe6e6a97b55f8fe4b00218ae11972902947ed058d8281d40cb1d6e6f3c  figures/accounting_summary.pdf
 a2a0284f9c0687ba6ff52b4a8077189d7a2ba21b63e5e77c4287639b6bbc23b4  figures/audit_protocol.pdf
 de3901b50dba906bc418660a4de8983ad311a0d4ed12b946e4d12b104435e4e2  figures/bootstrap_robustness.pdf
diff --git a/paper/results/generated_values.tex b/paper/results/generated_values.tex
index 555d9af..cc34141 100644
--- a/paper/results/generated_values.tex
+++ b/paper/results/generated_values.tex
@@ -4,7 +4,7 @@
 \newcommand{\PaperRequiredWarmupSessions}{274}
 \newcommand{\PaperBootstrapReplications}{5,000}
 \newcommand{\PaperInputDigest}{efb384a62157e56a0cd8065abf45c1ed07d90ec26c681e5d54d74fe4cb9c55e1}
-\newcommand{\PaperSourceTreeDigest}{df200735ec35013b252a3aa689752b9eae485c21abe7247bc590c444f18221b6}
+\newcommand{\PaperSourceTreeDigest}{044c219b82295b187b92a7044757289faf91757d276206fe6abdec5d52252e60}
 \newcommand{\PaperNetCAGR}{1.40\%}
 \newcommand{\PaperGrossCAGR}{3.17\%}
 \newcommand{\PaperCashCAGR}{2.50\%}
diff --git a/paper/results/manifest.json b/paper/results/manifest.json
index 49506ec..d6a84df 100644
--- a/paper/results/manifest.json
+++ b/paper/results/manifest.json
@@ -20,15 +20,81 @@
     "results/cost_sensitivity.csv": "25b9969cf8ecdabd19e7761ad2973252c9f5b9f994b6dc147595a2a112c24d88",
     "results/engine_comparison.csv": "ea0322dcd2aa4a1eb3b5996045c3a4b2ed25a85684a578ee788b013b99b643cd",
     "results/evaluation_contract.json": "77ee05ce64622ef9ba1bfbd7dae85c4f6fd44f07db0b21feaf6b3e0e418673e7",
-    "results/generated_values.tex": "f14a548364f13b40e75d44c2eb2c3aa0965668ac35a6acd30d22ef207abb1fdf",
+    "results/generated_values.tex": "98611c031c54382859488367dc1f5474a283014e3db2867f8210dab49ac85f25",
     "results/protocol_switches.csv": "20d5a0dc37e07a20c1f2772c8e9940ff6464ef1da98692edd5fdb06531bc2393",
     "results/return_decomposition.csv": "53a84aa9b91c92f0037e48dc09154807cbf003ccf0038f39da8406108bf709cc",
     "results/sharpe_uncertainty.csv": "aead873ced9b25c76c6944aa6e3ba0901ade65c784fa5cb8cc10ad1c6c01f136",
     "results/subperiods.csv": "9e6206fd38422640b4539e1207aade8821c2ee474af18e74268cb4bf69c249db",
+    "results/target_tape_hashes.json": "7eac818e4f37ed936561138c9a00f2038b6051723673a0162b609b2ae64bca03",
     "results/uncertainty.json": "6a6a0aa9a02def987f09de273d99cb28a6929cedf090d6245e2e1fa818d2c54a",
     "results/yearly_returns.csv": "53cfee7bb987c332b9db1766619b0eb085415e358724872a77ebd6c5f1ed6256"
   },
   "configuration_sha256": "20ab6c2850734e9c7a44eae17b4c3f7db4dbfda18d85cb9ecd9fcc729b475128",
+  "decision_streams": {
+    "path": "results/target_tape_hashes.json",
+    "variants": {
+      "full": {
+        "canonical_payload_sha256": "a5c103d7d53dbc538971a7f85674843e5ace76ea10e4e09cc061f6bc9a05aaf4",
+        "decision_count": 521,
+        "max_gross": 1.0,
+        "record_count": 3126,
+        "schema_version": 1,
+        "symbols": [
+          "GLD",
+          "QQQ",
+          "SPY",
+          "TLT",
+          "VGT",
+          "VIG"
+        ]
+      },
+      "no_regime": {
+        "canonical_payload_sha256": "cdbf2c1a4a500589ed0ee3b4954c2b85b9774bffd7fd046d77359a25c0908280",
+        "decision_count": 521,
+        "max_gross": 1.0,
+        "record_count": 3126,
+        "schema_version": 1,
+        "symbols": [
+          "GLD",
+          "QQQ",
+          "SPY",
+          "TLT",
+          "VGT",
+          "VIG"
+        ]
+      },
+      "no_vol_scaler": {
+        "canonical_payload_sha256": "faeb6bca8e7a467d670b0be6ad1cbd44192e468b159c1a30456afa4dcf58c7fa",
+        "decision_count": 521,
+        "max_gross": 1.0,
+        "record_count": 3126,
+        "schema_version": 1,
+        "symbols": [
+          "GLD",
+          "QQQ",
+          "SPY",
+          "TLT",
+          "VGT",
+          "VIG"
+        ]
+      },
+      "signal_only": {
+        "canonical_payload_sha256": "0746243609426143207f8c42a83026fb4f78c6564d893e57cab56831c7c18a69",
+        "decision_count": 521,
+        "max_gross": 1.0,
+        "record_count": 3126,
+        "schema_version": 1,
+        "symbols": [
+          "GLD",
+          "QQQ",
+          "SPY",
+          "TLT",
+          "VGT",
+          "VIG"
+        ]
+      }
+    }
+  },
   "design": {
     "all_in_cost_levels_bps": [
       0.0,
@@ -174,14 +240,14 @@
     "path": "results/evaluation_contract.json",
     "schema_version": 1
   },
-  "generated_at": "2026-06-16T22:00:41Z",
+  "generated_at": "2026-06-18T22:06:35Z",
   "generator": {
     "dependency_lock": {
       "path": "poetry.lock",
-      "sha256": "fa8904674c2b24f3141acae607fd0a49d97125bc109c18def9fccbb2256c8de0"
+      "sha256": "abce2cd132651851d81034dea085d7dbfbe5e44321f2e13997db69f46d7f080b"
     },
     "git": {
-      "source_commit": "d185c44928af865882c4d830066432dc97eb9972",
+      "source_commit": "ef6e7e3414aed0ae9b77d50748a22823529486af",
       "worktree_clean_at_start": true
     },
     "packages": {
@@ -195,12 +261,12 @@
     "path": "scripts/run_paper_experiments.py",
     "platform": "macOS-26.4-arm64-arm-64bit-Mach-O",
     "python": "3.14.4",
-    "script_sha256": "2f7c00e1d387d7efde57c0264d3cd702176af2fa94e5f5646309b0641ddb34bc",
+    "script_sha256": "922bf3c414e0eadca5c13ee831528347dd83a0cd174efbdf15a987b511845de5",
     "source_tree": {
-      "file_count": 109,
+      "file_count": 110,
       "files": {
-        "poetry.lock": "fa8904674c2b24f3141acae607fd0a49d97125bc109c18def9fccbb2256c8de0",
-        "pyproject.toml": "175abd743d38f6b6c1cc8ae42471d62d83c86beb245d93fea231c666f51b2643",
+        "poetry.lock": "abce2cd132651851d81034dea085d7dbfbe5e44321f2e13997db69f46d7f080b",
+        "pyproject.toml": "eaeeb454c28bf7f6d9e530002bb7e88624b56b6c3e1fcb71e6414045cb9c42a0",
         "quantcortex/__init__.py": "14bf1ebdacd054c3738e4704d33da6709a39206463df8b8ced5376da342c4036",
         "quantcortex/alpha/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
         "quantcortex/alpha/factors/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
@@ -224,7 +290,7 @@
         "quantcortex/alpha/validation/alphalens_report.py": "8463289117add78a576e37a5b99a539be5c808fca86100e26deb29e8060aa60d",
         "quantcortex/alpha/validation/factor_decay.py": "9e6e049165f014db2122d9ed57415e45e48142b475dfb7221247d447d9a50397",
         "quantcortex/backtest/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
-        "quantcortex/backtest/conformance.py": "e169cc67cac9eb56d8da4e4a0da1e257bb6cb8d1745baca681f50fab6c62fc57",
+        "quantcortex/backtest/conformance.py": "22f9b18101f0d9adcae98297e395c1ab6a90b18e411f24e7456f1d166b7df9b5",
         "quantcortex/backtest/costs/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
         "quantcortex/backtest/costs/transaction_costs.py": "48200419f37004c58df6f89041e31516d9ad98f66c22f75d11a0b9342474d5c7",
         "quantcortex/backtest/engines/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
@@ -245,7 +311,7 @@
         "quantcortex/backtest/validation/multiple_testing.py": "6b68f6f150dcc8fb90213dcf769307454c0c506e99666a533d5d07ae3ccb0b5d",
         "quantcortex/backtest/validation/survivorship_check.py": "5b772bd6b40ffe44f590ec0ddb00b7e3dc61a1aba9cae8831227325f4f53beff",
         "quantcortex/data/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
-        "quantcortex/data/local_csv.py": "30c11c3200530810b3345f8e29fdf966f35671e30c94fc5bb1b2db656e6df79b",
+        "quantcortex/data/local_csv.py": "6e5f3d24b1477ce122a1e80899e841fffd6d857034189f7d8605705f52c6839c",
         "quantcortex/data/processors/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
         "quantcortex/data/processors/adjustments.py": "ae2d4d0362e3484b89b5a5c57a7fd6320973b208ee0cd5a6be9a3f29d2544e2c",
         "quantcortex/data/processors/calendar.py": "e591bccc3e7f632501b068e7abec1ece182d0ef930589381776fe89158d1df51",
@@ -305,11 +371,12 @@
         "quantcortex/timing/kama.py": "ffb848ba93041af1e04f2b7896ea7a8f33ac7e5e3e53db1e44f53af111829eda",
         "quantcortex/timing/tsmom.py": "c78ac313268091f45e0e589067f12e07e8a340fa243b8fae2853778ef4ba8ea9",
         "quantcortex/timing/vix_scaler.py": "a3667424e5573fb289e63c26c69da6a68d6c943742359f0466d29b25c56e3686",
-        "schemas/canonical_target_tape.schema.json": "81acbd7e03163ea0dc28f7b1a4bd2a73a4bbeef85dbdbc70aee01aa100c4e7e5",
+        "schemas/canonical_target_tape.schema.json": "4f1c0bf6d5360305d2982adea78de3f61c4bc1ebae9207cb2ba2bd4379b43d44",
         "schemas/evaluation_contract.schema.json": "970f24f587e669925306625d12c5a84dffd03ff5b222a59905849b2fa222784f",
-        "scripts/run_paper_experiments.py": "2f7c00e1d387d7efde57c0264d3cd702176af2fa94e5f5646309b0641ddb34bc"
+        "scripts/release_paper_artifacts.sh": "cf3de9434ab3991598e0c8d1d2c9346fde425fdf1eb792a6941fede0c66342e9",
+        "scripts/run_paper_experiments.py": "922bf3c414e0eadca5c13ee831528347dd83a0cd174efbdf15a987b511845de5"
       },
-      "sha256": "df200735ec35013b252a3aa689752b9eae485c21abe7247bc590c444f18221b6"
+      "sha256": "044c219b82295b187b92a7044757289faf91757d276206fe6abdec5d52252e60"
     },
     "threadpools": [
       {
@@ -321,7 +388,7 @@
       }
     ]
   },
-  "schema_version": 4,
+  "schema_version": 5,
   "source": {
     "adjustment_method": "yfinance adjusted close with auto_adjust=False",
     "cash_proxy": "SHV",
diff --git a/paper/results/target_tape_hashes.json b/paper/results/target_tape_hashes.json
new file mode 100644
index 0000000..17dff2a
--- /dev/null
+++ b/paper/results/target_tape_hashes.json
@@ -0,0 +1,62 @@
+{
+  "full": {
+    "canonical_payload_sha256": "a5c103d7d53dbc538971a7f85674843e5ace76ea10e4e09cc061f6bc9a05aaf4",
+    "decision_count": 521,
+    "max_gross": 1.0,
+    "record_count": 3126,
+    "schema_version": 1,
+    "symbols": [
+      "GLD",
+      "QQQ",
+      "SPY",
+      "TLT",
+      "VGT",
+      "VIG"
+    ]
+  },
+  "no_regime": {
+    "canonical_payload_sha256": "cdbf2c1a4a500589ed0ee3b4954c2b85b9774bffd7fd046d77359a25c0908280",
+    "decision_count": 521,
+    "max_gross": 1.0,
+    "record_count": 3126,
+    "schema_version": 1,
+    "symbols": [
+      "GLD",
+      "QQQ",
+      "SPY",
+      "TLT",
+      "VGT",
+      "VIG"
+    ]
+  },
+  "no_vol_scaler": {
+    "canonical_payload_sha256": "faeb6bca8e7a467d670b0be6ad1cbd44192e468b159c1a30456afa4dcf58c7fa",
+    "decision_count": 521,
+    "max_gross": 1.0,
+    "record_count": 3126,
+    "schema_version": 1,
+    "symbols": [
+      "GLD",
+      "QQQ",
+      "SPY",
+      "TLT",
+      "VGT",
+      "VIG"
+    ]
+  },
+  "signal_only": {
+    "canonical_payload_sha256": "0746243609426143207f8c42a83026fb4f78c6564d893e57cab56831c7c18a69",
+    "decision_count": 521,
+    "max_gross": 1.0,
+    "record_count": 3126,
+    "schema_version": 1,
+    "symbols": [
+      "GLD",
+      "QQQ",
+      "SPY",
+      "TLT",
+      "VGT",
+      "VIG"
+    ]
+  }
+}
diff --git a/poetry.lock b/poetry.lock
index ac8e3df..b0cc87e 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -403,7 +403,7 @@ version = "1.4.0"
 description = "Better dates & times for Python"
 optional = false
 python-versions = ">=3.8"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "arrow-1.4.0-py3-none-any.whl", hash = "sha256:749f0769958ebdc79c173ff0b0670d59051a535fa26e8eba02953dc19eb43205"},
     {file = "arrow-1.4.0.tar.gz", hash = "sha256:ed0cc050e98001b8779e84d461b0098c4ac597e88704a655582b21d116e526d7"},
@@ -464,7 +464,7 @@ version = "26.1.0"
 description = "Classes Without Boilerplate"
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "dev"]
+groups = ["main", "dev", "test"]
 files = [
     {file = "attrs-26.1.0-py3-none-any.whl", hash = "sha256:c647aa4a12dfbad9333ca4e71fe62ddc36f4e63b2d260a37a8b83d2f043ac309"},
     {file = "attrs-26.1.0.tar.gz", hash = "sha256:d03ceb89cb322a8fd706d4fb91940737b6642aa36998fe130a9bc96c985eff32"},
@@ -1632,7 +1632,7 @@ version = "1.5.1"
 description = "Validates fully-qualified domain names against RFC 1123, so that they are acceptable to modern bowsers"
 optional = false
 python-versions = ">=2.7, !=3.0, !=3.1, !=3.2, !=3.3, !=3.4, <4"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "fqdn-1.5.1-py3-none-any.whl", hash = "sha256:3a179af3761e4df6eb2e026ff9e1a3033d3587bf980a0b1b2e1e5d08d7358014"},
     {file = "fqdn-1.5.1.tar.gz", hash = "sha256:105ed3677e767fb5ca086a0c1f4bb66ebc3c100be518f0e0d755d9eae164d89f"},
@@ -2262,7 +2262,7 @@ version = "3.18"
 description = "Internationalized Domain Names in Applications (IDNA)"
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "dev"]
+groups = ["main", "dev", "test"]
 files = [
     {file = "idna-3.18-py3-none-any.whl", hash = "sha256:7f952cbe720b688055e3f87de14f5c3e5fdaa8bc3928985c4077ca689de849a2"},
     {file = "idna-3.18.tar.gz", hash = "sha256:ffb385a7e039654cef1ab9ef32c6fafe283c0c0467bba1d9029738ce4a14a848"},
@@ -2395,7 +2395,7 @@ version = "20.11.0"
 description = "Operations with ISO 8601 durations"
 optional = false
 python-versions = ">=3.7"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "isoduration-20.11.0-py3-none-any.whl", hash = "sha256:b2904c2a4228c3d44f409c8ae8e2370eb21a26f7ac2ec5446df141dde3452042"},
     {file = "isoduration-20.11.0.tar.gz", hash = "sha256:ac2f9015137935279eac671f94f89eb00584f940f5dc49462a0c4ee692ba1bd9"},
@@ -2472,7 +2472,7 @@ version = "3.1.1"
 description = "Identify specific nodes in a JSON document (RFC 6901) "
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "jsonpointer-3.1.1-py3-none-any.whl", hash = "sha256:8ff8b95779d071ba472cf5bc913028df06031797532f08a7d5b602d8b2a488ca"},
     {file = "jsonpointer-3.1.1.tar.gz", hash = "sha256:0b801c7db33a904024f6004d526dcc53bbb8a4a0f4e32bfd10beadf60adf1900"},
@@ -2484,7 +2484,7 @@ version = "4.26.0"
 description = "An implementation of JSON Schema validation for Python"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "jsonschema-4.26.0-py3-none-any.whl", hash = "sha256:d489f15263b8d200f8387e64b4c3a75f06629559fb73deb8fdfb525f2dab50ce"},
     {file = "jsonschema-4.26.0.tar.gz", hash = "sha256:0c26707e2efad8aa1bfc5b7ce170f3fccc2e4918ff85989ba9ffa9facb2be326"},
@@ -2515,7 +2515,7 @@ version = "2025.9.1"
 description = "The JSON Schema meta-schemas and vocabularies, exposed as a Registry"
 optional = false
 python-versions = ">=3.9"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "jsonschema_specifications-2025.9.1-py3-none-any.whl", hash = "sha256:98802fee3a11ee76ecaca44429fda8a41bff98b00a0f2838151b113f210cc6fe"},
     {file = "jsonschema_specifications-2025.9.1.tar.gz", hash = "sha256:b540987f239e745613c7a9176f3edb72b832a4ac465cf02712288397832b5e8d"},
@@ -2930,7 +2930,7 @@ version = "1.3.1"
 description = "a modern parsing library"
 optional = false
 python-versions = ">=3.8"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "lark-1.3.1-py3-none-any.whl", hash = "sha256:c629b661023a014c37da873b4ff58a817398d12635d3bbb2c5a03be7fe5d1e12"},
     {file = "lark-1.3.1.tar.gz", hash = "sha256:b426a7a6d6d53189d318f2b6236ab5d6429eaf09259f1ca33eb716eed10d2905"},
@@ -5370,7 +5370,7 @@ version = "2.9.0.post0"
 description = "Extensions to the standard Python datetime module"
 optional = false
 python-versions = "!=3.0.*,!=3.1.*,!=3.2.*,>=2.7"
-groups = ["main", "dev"]
+groups = ["main", "dev", "test"]
 files = [
     {file = "python-dateutil-2.9.0.post0.tar.gz", hash = "sha256:37dd54208da7e1cd875388217d5e00ebd4179249f90fb72437e91a35459a0ad3"},
     {file = "python_dateutil-2.9.0.post0-py2.py3-none-any.whl", hash = "sha256:a8b2bc7bffae282281c8140a97d3aa9c14da0b136dfe83f850eea9a5f7470427"},
@@ -5652,7 +5652,7 @@ version = "0.37.0"
 description = "JSON Referencing + Python"
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "referencing-0.37.0-py3-none-any.whl", hash = "sha256:381329a9f99628c9069361716891d34ad94af76e461dcb0335825aecc7692231"},
     {file = "referencing-0.37.0.tar.gz", hash = "sha256:44aefc3142c5b842538163acb373e24cce6632bd54bdb01b21ad5863489f50d8"},
@@ -5817,7 +5817,7 @@ version = "0.1.4"
 description = "A pure python RFC3339 validator"
 optional = false
 python-versions = ">=2.7, !=3.0.*, !=3.1.*, !=3.2.*, !=3.3.*, !=3.4.*"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "rfc3339_validator-0.1.4-py2.py3-none-any.whl", hash = "sha256:24f6ec1eda14ef823da9e36ec7113124b39c04d50a4d3d3a3c2859577e7791fa"},
     {file = "rfc3339_validator-0.1.4.tar.gz", hash = "sha256:138a2abdf93304ad60530167e51d2dfb9549521a836871b88d7f4695d0022f6b"},
@@ -5832,7 +5832,7 @@ version = "0.1.1"
 description = "Pure python rfc3986 validator"
 optional = false
 python-versions = ">=2.7, !=3.0.*, !=3.1.*, !=3.2.*, !=3.3.*, !=3.4.*"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "rfc3986_validator-0.1.1-py2.py3-none-any.whl", hash = "sha256:2f235c432ef459970b4306369336b9d5dbdda31b510ca1e327636e01f528bfa9"},
     {file = "rfc3986_validator-0.1.1.tar.gz", hash = "sha256:3d44bde7921b3b9ec3ae4e3adca370438eccebc676456449b145d533b240d055"},
@@ -5844,7 +5844,7 @@ version = "1.1.0"
 description = "Helper functions to syntactically validate strings according to RFC 3987."
 optional = false
 python-versions = ">=3.9"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "rfc3987_syntax-1.1.0-py3-none-any.whl", hash = "sha256:6c3d97604e4c5ce9f714898e05401a0445a641cfa276432b0a648c80856f6a3f"},
     {file = "rfc3987_syntax-1.1.0.tar.gz", hash = "sha256:717a62cbf33cffdd16dfa3a497d81ce48a660ea691b1ddd7be710c22f00b4a0d"},
@@ -5882,7 +5882,7 @@ version = "2026.5.1"
 description = "Python bindings to Rust's persistent data structures (rpds)"
 optional = false
 python-versions = ">=3.11"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "rpds_py-2026.5.1-cp311-cp311-macosx_10_12_x86_64.whl", hash = "sha256:3397a5ed7174dc2786bb214030232fc36fe8e5584fec43a9952cc542b1a12036"},
     {file = "rpds_py-2026.5.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:99ab6ba7bfa2cb0f96a04e3652355bf04e3f51aceb1e943b8541dab7ba4828cc"},
@@ -6280,7 +6280,7 @@ version = "1.17.0"
 description = "Python 2 and 3 compatibility utilities"
 optional = false
 python-versions = "!=3.0.*,!=3.1.*,!=3.2.*,>=2.7"
-groups = ["main", "dev"]
+groups = ["main", "dev", "test"]
 files = [
     {file = "six-1.17.0-py2.py3-none-any.whl", hash = "sha256:4721f391ed90541fddacab5acf947aa0d3dc7d27b2e1e8eda2be8970586c3274"},
     {file = "six-1.17.0.tar.gz", hash = "sha256:ff70335d468e7eb6ec65b95b99d3a2836546063f63acc5171de367e834932a81"},
@@ -6846,12 +6846,12 @@ version = "4.15.0"
 description = "Backported and Experimental Type Hints for Python 3.9+"
 optional = false
 python-versions = ">=3.9"
-groups = ["main", "dev"]
+groups = ["main", "dev", "test"]
 files = [
     {file = "typing_extensions-4.15.0-py3-none-any.whl", hash = "sha256:f0fa19c6845758ab08074a0cfa8b7aecb71c999ca73d62883bc25cc018c4e548"},
     {file = "typing_extensions-4.15.0.tar.gz", hash = "sha256:0cea48d173cc12fa28ecabc3b837ea3cf6f38c6d1136f85cbaaf598984861466"},
 ]
-markers = {main = "python_version < \"3.14\" and (extra == \"brokers\" or extra == \"all\" or extra == \"nlp\" or extra == \"rl\" or extra == \"providers\" or extra == \"storage\") or extra == \"rl\" or extra == \"all\" or extra == \"nlp\" or extra == \"brokers\" or extra == \"providers\" or extra == \"storage\""}
+markers = {main = "python_version < \"3.14\" and (extra == \"brokers\" or extra == \"all\" or extra == \"nlp\" or extra == \"rl\" or extra == \"providers\" or extra == \"storage\") or extra == \"rl\" or extra == \"all\" or extra == \"nlp\" or extra == \"brokers\" or extra == \"providers\" or extra == \"storage\"", test = "python_version < \"3.13\""}
 
 [[package]]
 name = "typing-inspection"
@@ -6875,7 +6875,7 @@ version = "2025.3"
 description = "Provider of IANA time zone data"
 optional = false
 python-versions = ">=2"
-groups = ["main", "dev"]
+groups = ["main", "dev", "test"]
 files = [
     {file = "tzdata-2025.3-py2.py3-none-any.whl", hash = "sha256:06a47e5700f3081aab02b2e513160914ff0694bce9947d6b76ebd6bf57cfc5d1"},
     {file = "tzdata-2025.3.tar.gz", hash = "sha256:de39c2ca5dc7b0344f2eba86f49d614019d29f060fc4ebc8a417896a620b56a7"},
@@ -6888,7 +6888,7 @@ version = "1.3.0"
 description = "RFC 6570 URI Template Processor"
 optional = false
 python-versions = ">=3.7"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "uri-template-1.3.0.tar.gz", hash = "sha256:0e00f8eb65e18c7de20d595a14336e9f337ead580c70934141624b6d1ffdacc7"},
     {file = "uri_template-1.3.0-py3-none-any.whl", hash = "sha256:a44a133ea12d44a0c0f06d7d42a52d71282e77e2f937d8abd5655b8d56fc1363"},
@@ -6934,7 +6934,7 @@ version = "25.10.0"
 description = "A library for working with the color formats defined by HTML and CSS."
 optional = false
 python-versions = ">=3.10"
-groups = ["dev"]
+groups = ["dev", "test"]
 files = [
     {file = "webcolors-25.10.0-py3-none-any.whl", hash = "sha256:032c727334856fc0b968f63daa252a1ac93d33db2f5267756623c210e57a4f1d"},
     {file = "webcolors-25.10.0.tar.gz", hash = "sha256:62abae86504f66d0f6364c2a8520de4a0c47b80c03fc3a5f1815fedbef7c19bf"},
@@ -7264,4 +7264,4 @@ storage = ["psycopg2-binary", "redis", "sqlalchemy"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.11,<3.15"
-content-hash = "fdf69ed71f0e77bc4e134fd0c617b6ad4ecb2bc28e3b34547f8513519bcbf3ed"
+content-hash = "20db487a78d05dfdbc61f18045e883b20a6a82fc2a028ed30bd7902bf816f6ff"
diff --git a/pyproject.toml b/pyproject.toml
index 72c7f96..c67df2e 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -86,6 +86,7 @@ packages = [
 poetry-plugin-export = ">=1.10,<1.11"
 
 [tool.poetry.group.test.dependencies]
+jsonschema = { version = ">=4.20,<5", extras = ["format-nongpl"] }
 packaging = ">=24"
 pytest = ">=8.0"
 pytest-cov = ">=4.1"
diff --git a/quantcortex/backtest/conformance.py b/quantcortex/backtest/conformance.py
index 1d36a1c..898e48c 100644
--- a/quantcortex/backtest/conformance.py
+++ b/quantcortex/backtest/conformance.py
@@ -8,7 +8,9 @@
 
 from __future__ import annotations
 
+import re
 from collections.abc import Mapping, Sequence
+from datetime import datetime
 
 import numpy as np
 import pandas as pd
@@ -19,6 +21,69 @@
     "target_weight",
 )
 TARGET_TAPE_SCHEMA_VERSION = 1
+_RFC3339_TIMESTAMP = re.compile(
+    r"^\d{4}-\d{2}-\d{2}T(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d"
+    r"(?:\.\d+)?(?:Z|[+-](?:[01]\d|2[0-3]):[0-5]\d)$"
+)
+
+
+def _validated_payload_timestamp(value: object) -> str:
+    """Validate the RFC 3339 subset emitted by the canonical serializer."""
+    if not isinstance(value, str):
+        raise TypeError(
+            "target-tape decision_timestamp must be an RFC 3339 string"
+        )
+    if _RFC3339_TIMESTAMP.fullmatch(value) is None:
+        raise ValueError(
+            "target-tape decision_timestamp must be an RFC 3339 date-time"
+        )
+    candidate = value[:-1] + "+00:00" if value.endswith("Z") else value
+    try:
+        parsed = datetime.fromisoformat(candidate)
+    except ValueError as exc:
+        raise ValueError(
+            "target-tape decision_timestamp must be an RFC 3339 date-time"
+        ) from exc
+    if parsed.tzinfo is None or parsed.utcoffset() is None:
+        raise ValueError(
+            "target-tape decision_timestamp must include a UTC offset"
+        )
+    return value
+
+
+def _validated_payload_number(value: object, *, name: str) -> float:
+    """Accept only JSON number primitives, excluding booleans and strings."""
+    if type(value) not in (int, float):
+        raise TypeError(f"{name} must be a JSON number")
+    try:
+        parsed = float(value)
+    except OverflowError as exc:
+        raise ValueError(f"{name} must be finite binary64") from exc
+    if not np.isfinite(parsed):
+        raise ValueError(f"{name} must be finite binary64")
+    return parsed
+
+
+def _validated_payload_schema_version(value: object) -> int:
+    """Normalize JSON numbers that represent an integral schema version."""
+    parsed = _validated_payload_number(value, name="target-tape schema_version")
+    if not parsed.is_integer():
+        raise TypeError("target-tape schema_version must be an integer")
+    version = int(parsed)
+    if version != TARGET_TAPE_SCHEMA_VERSION:
+        raise ValueError("unsupported target-tape schema version")
+    return version
+
+
+def _validated_payload_symbol(value: object, *, name: str) -> str:
+    """Validate a canonical wire symbol without silently normalizing it."""
+    if not isinstance(value, str):
+        raise TypeError(f"{name} must be a string")
+    if not value or value != value.strip() or "\r" in value or "\n" in value:
+        raise ValueError(
+            f"{name} must be non-empty with no outer whitespace or line breaks"
+        )
+    return value
 
 
 def _validated_max_gross(value: object) -> float:
@@ -61,6 +126,7 @@ def validate_target_tape(
     try:
         timestamps = pd.to_datetime(
             normalized["decision_timestamp"],
+            format="mixed",
             utc=True,
             errors="raise",
         )
@@ -88,7 +154,7 @@ def validate_target_tape(
         raise ValueError("target_weight values must be finite")
     if (weights < -1e-12).any():
         raise ValueError("target_weight values must be long-only")
-    normalized.loc[normalized["target_weight"].abs() < 1e-15, "target_weight"] = 0.0
+    normalized.loc[normalized["target_weight"] < 0.0, "target_weight"] = 0.0
 
     duplicate = normalized.duplicated(
         subset=["decision_timestamp", "symbol"],
@@ -154,7 +220,12 @@ def target_tape_to_weights(
     )
     weights.columns.name = None
     weights.index.name = None
-    return weights.sort_index().sort_index(axis=1)
+    column_order = (
+        sorted(normalized["symbol"].unique())
+        if expected_symbols is None
+        else [symbol.strip() for symbol in expected_symbols]
+    )
+    return weights.sort_index().reindex(columns=column_order)
 
 
 def weights_to_target_tape(
@@ -234,11 +305,14 @@ def target_tape_from_payload(payload: Mapping[str, object]) -> pd.DataFrame:
             "target-tape payload keys must be exactly "
             + ", ".join(sorted(required_keys))
         )
-    if payload["schema_version"] != TARGET_TAPE_SCHEMA_VERSION:
-        raise ValueError("unsupported target-tape schema version")
+    _validated_payload_schema_version(payload["schema_version"])
     symbols = payload["symbols"]
     if not isinstance(symbols, list):
         raise TypeError("target-tape symbols must be a list")
+    symbols = [
+        _validated_payload_symbol(symbol, name="target-tape symbol")
+        for symbol in symbols
+    ]
     records = payload["records"]
     if not isinstance(records, list) or not records:
         raise ValueError("target-tape records must be a non-empty list")
@@ -249,10 +323,33 @@ def target_tape_from_payload(payload: Mapping[str, object]) -> pd.DataFrame:
             "target-tape record keys must be exactly "
             + ", ".join(TARGET_TAPE_COLUMNS)
         )
-    tape = pd.DataFrame(records, columns=TARGET_TAPE_COLUMNS)
+    normalized_records = []
+    for record in records:
+        timestamp = _validated_payload_timestamp(record["decision_timestamp"])
+        symbol = _validated_payload_symbol(
+            record["symbol"],
+            name="target-tape record symbol",
+        )
+        target_weight = _validated_payload_number(
+            record["target_weight"],
+            name="target-tape target_weight",
+        )
+        if target_weight < 0.0:
+            raise ValueError("target-tape target_weight must be non-negative")
+        normalized_records.append(
+            {
+                "decision_timestamp": timestamp,
+                "symbol": symbol,
+                "target_weight": target_weight,
+            }
+        )
+    max_gross = _validated_payload_number(
+        payload["max_gross"],
+        name="target-tape max_gross",
+    )
     return validate_target_tape(
-        tape,
-        max_gross=_validated_max_gross(payload["max_gross"]),
+        pd.DataFrame(normalized_records, columns=TARGET_TAPE_COLUMNS),
+        max_gross=_validated_max_gross(max_gross),
         expected_symbols=symbols,
     )
 
diff --git a/quantcortex/data/local_csv.py b/quantcortex/data/local_csv.py
index bb6084c..58c8a63 100644
--- a/quantcortex/data/local_csv.py
+++ b/quantcortex/data/local_csv.py
@@ -95,12 +95,16 @@ def load_price_matrix(
     start: str | None = None,
     end: str | None = None,
     max_ffill: int | None = 5,
+    *,
+    require_complete: bool = False,
 ) -> pd.DataFrame:
     """Load a wide adjusted-close CSV indexed by a required ``date`` column.
 
     Missing prices are forward-filled for at most ``max_ffill`` rows. Set it to
     ``None`` to disable filling; unlimited filling is intentionally unsupported
-    because it can keep stale or delisted assets alive indefinitely.
+    because it can keep stale or delisted assets alive indefinitely. When
+    ``require_complete`` is true, any missing value that remains in the requested
+    date window raises instead of silently removing that row.
     """
     resolved, frame = _read_dated_csv(path)
     frame.columns = [str(column).strip() for column in frame.columns]
@@ -126,6 +130,8 @@ def load_price_matrix(
     frame = _numeric(frame, resolved)
     if (frame <= 0).any(axis=None):
         raise LocalDataError(f"prices must be strictly positive in {resolved}")
+    if not isinstance(require_complete, bool):
+        raise LocalDataError("require_complete must be a boolean")
 
     if max_ffill is not None:
         if (
@@ -135,7 +141,20 @@ def load_price_matrix(
         ):
             raise LocalDataError("max_ffill must be a non-negative integer or None")
         frame = frame.ffill(limit=int(max_ffill)) if max_ffill > 0 else frame
-    frame = _slice(frame, start, end).dropna(how="any")
+    frame = _slice(frame, start, end)
+    if require_complete and frame.isna().any(axis=None):
+        missing = frame.isna().stack()
+        locations = [
+            f"{timestamp.date().isoformat()}:{symbol}"
+            for (timestamp, symbol), is_missing in missing.items()
+            if is_missing
+        ]
+        preview = ", ".join(locations[:3])
+        raise LocalDataError(
+            "price matrix contains missing required observations"
+            + (f"; first missing values: {preview}" if preview else "")
+        )
+    frame = frame.dropna(how="any")
     if frame.empty:
         raise LocalDataError("no complete price rows remain after forward-filling")
     return frame
diff --git a/requirements/test.lock b/requirements/test.lock
index 0d7d335..0f53e4f 100644
--- a/requirements/test.lock
+++ b/requirements/test.lock
@@ -1,11 +1,20 @@
+arrow==1.4.0 ; python_version >= "3.11" and python_version < "3.15"
+attrs==26.1.0 ; python_version >= "3.11" and python_version < "3.15"
 colorama==0.4.6 ; python_version >= "3.11" and python_version < "3.15" and (platform_system == "Windows" or sys_platform == "win32")
 contourpy==1.3.3 ; python_version >= "3.11" and python_version < "3.15"
 coverage==7.14.1 ; python_version >= "3.11" and python_version < "3.15"
 cycler==0.12.1 ; python_version >= "3.11" and python_version < "3.15"
 fonttools==4.63.0 ; python_version >= "3.11" and python_version < "3.15"
+fqdn==1.5.1 ; python_version >= "3.11" and python_version < "3.15"
+idna==3.18 ; python_version >= "3.11" and python_version < "3.15"
 iniconfig==2.3.0 ; python_version >= "3.11" and python_version < "3.15"
+isoduration==20.11.0 ; python_version >= "3.11" and python_version < "3.15"
 joblib==1.5.3 ; python_version >= "3.11" and python_version < "3.15"
+jsonpointer==3.1.1 ; python_version >= "3.11" and python_version < "3.15"
+jsonschema-specifications==2025.9.1 ; python_version >= "3.11" and python_version < "3.15"
+jsonschema==4.26.0 ; python_version >= "3.11" and python_version < "3.15"
 kiwisolver==1.5.0 ; python_version >= "3.11" and python_version < "3.15"
+lark==1.3.1 ; python_version >= "3.11" and python_version < "3.15"
 matplotlib==3.11.0 ; python_version >= "3.11" and python_version < "3.15"
 narwhals==2.22.1 ; python_version >= "3.11" and python_version < "3.15"
 numpy==2.4.6 ; python_version >= "3.11" and python_version < "3.15"
@@ -19,9 +28,17 @@ pyparsing==3.3.2 ; python_version >= "3.11" and python_version < "3.15"
 pytest-cov==7.1.0 ; python_version >= "3.11" and python_version < "3.15"
 pytest==9.1.0 ; python_version >= "3.11" and python_version < "3.15"
 python-dateutil==2.9.0.post0 ; python_version >= "3.11" and python_version < "3.15"
+referencing==0.37.0 ; python_version >= "3.11" and python_version < "3.15"
+rfc3339-validator==0.1.4 ; python_version >= "3.11" and python_version < "3.15"
+rfc3986-validator==0.1.1 ; python_version >= "3.11" and python_version < "3.15"
+rfc3987-syntax==1.1.0 ; python_version >= "3.11" and python_version < "3.15"
+rpds-py==2026.5.1 ; python_version >= "3.11" and python_version < "3.15"
 ruff==0.15.17 ; python_version >= "3.11" and python_version < "3.15"
 scikit-learn==1.9.0 ; python_version >= "3.11" and python_version < "3.15"
 scipy==1.17.1 ; python_version >= "3.11" and python_version < "3.15"
 six==1.17.0 ; python_version >= "3.11" and python_version < "3.15"
 threadpoolctl==3.6.0 ; python_version >= "3.11" and python_version < "3.15"
-tzdata==2025.3 ; python_version >= "3.11" and python_version < "3.15" and (sys_platform == "win32" or sys_platform == "emscripten")
+typing-extensions==4.15.0 ; python_version >= "3.11" and python_version < "3.14"
+tzdata==2025.3 ; python_version >= "3.11" and python_version < "3.15"
+uri-template==1.3.0 ; python_version >= "3.11" and python_version < "3.15"
+webcolors==25.10.0 ; python_version >= "3.11" and python_version < "3.15"
diff --git a/schemas/canonical_target_tape.schema.json b/schemas/canonical_target_tape.schema.json
index 03ec18d..2c01426 100644
--- a/schemas/canonical_target_tape.schema.json
+++ b/schemas/canonical_target_tape.schema.json
@@ -7,6 +7,7 @@
   "required": ["schema_version", "symbols", "max_gross", "records"],
   "properties": {
     "schema_version": {
+      "type": "integer",
       "const": 1
     },
     "symbols": {
@@ -16,12 +17,13 @@
       "items": {
         "type": "string",
         "minLength": 1,
-        "pattern": ".*\\S.*"
+        "pattern": "^(?:\\S|\\S.*\\S)$"
       }
     },
     "max_gross": {
       "type": "number",
-      "exclusiveMinimum": 0.0
+      "exclusiveMinimum": 0.0,
+      "maximum": 1.7976931348623157e308
     },
     "records": {
       "type": "array",
@@ -37,15 +39,18 @@
         "properties": {
           "decision_timestamp": {
             "type": "string",
-            "format": "date-time"
+            "format": "date-time",
+            "pattern": "^\\d{4}-\\d{2}-\\d{2}T(?:[01]\\d|2[0-3]):[0-5]\\d:[0-5]\\d(?:\\.\\d+)?(?:Z|[+-](?:[01]\\d|2[0-3]):[0-5]\\d)$"
           },
           "symbol": {
             "type": "string",
-            "minLength": 1
+            "minLength": 1,
+            "pattern": "^(?:\\S|\\S.*\\S)$"
           },
           "target_weight": {
             "type": "number",
-            "minimum": 0.0
+            "minimum": 0.0,
+            "maximum": 1.7976931348623157e308
           }
         }
       }
diff --git a/scripts/release_paper_artifacts.sh b/scripts/release_paper_artifacts.sh
index 7ca4b2f..f2f8a96 100755
--- a/scripts/release_paper_artifacts.sh
+++ b/scripts/release_paper_artifacts.sh
@@ -26,8 +26,93 @@ if [[ ! -f "${prices_csv}" ]]; then
   exit 1
 fi
 
-source_commit="$(git -C "${repo_root}" rev-parse HEAD)"
-generated_at="${QUANTCORTEX_GENERATED_AT:-$(date -u '+%Y-%m-%dT%H:%M:%SZ')}"
+reviewed_manifest="${repo_root}/paper/results/manifest.json"
+if [[ ! -f "${reviewed_manifest}" ]]; then
+  printf '%s\n' \
+    "reviewed paper manifest not found: ${reviewed_manifest}" >&2
+  exit 1
+fi
+expected_input_digest="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["source"]["input_sha256"])' \
+    "${reviewed_manifest}"
+)"
+actual_input_digest="$(
+  "${python_bin}" -c \
+    'import hashlib, pathlib, sys; print(hashlib.sha256(pathlib.Path(sys.argv[1]).read_bytes()).hexdigest())' \
+    "${prices_csv}"
+)"
+if [[ "${actual_input_digest}" != "${expected_input_digest}" ]]; then
+  printf '%s\n' \
+    "price matrix digest does not match the reviewed experiment" \
+    "expected: ${expected_input_digest}" \
+    "actual:   ${actual_input_digest}" >&2
+  exit 1
+fi
+case "${prices_csv}" in
+  "${repo_root}"/*)
+    relative_prices="${prices_csv#${repo_root}/}"
+    if git -C "${repo_root}" ls-files --error-unmatch -- "${relative_prices}" \
+      >/dev/null 2>&1; then
+      printf '%s\n' \
+        "reviewed raw price input must not be tracked by Git: ${relative_prices}" >&2
+      exit 1
+    else
+      tracking_status=$?
+      if [[ "${tracking_status}" -ne 1 ]]; then
+        printf '%s\n' \
+          "could not determine whether the raw price input is tracked" >&2
+        exit 1
+      fi
+    fi
+    ;;
+esac
+
+current_commit="$(git -C "${repo_root}" rev-parse HEAD)"
+reviewed_source_commit="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["generator"]["git"]["source_commit"])' \
+    "${reviewed_manifest}"
+)"
+reviewed_generated_at="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["generated_at"])' \
+    "${reviewed_manifest}"
+)"
+if [[ -n "${QUANTCORTEX_GENERATED_AT:-}" ]]; then
+  source_commit="${current_commit}"
+  generated_at="${QUANTCORTEX_GENERATED_AT}"
+else
+  if ! git -C "${repo_root}" cat-file -e "${reviewed_source_commit}^{commit}" \
+    >/dev/null 2>&1; then
+    printf '%s\n' "reviewed source commit is unavailable" >&2
+    exit 1
+  fi
+  release_source_paths=(
+    quantcortex
+    schemas
+    pyproject.toml
+    poetry.lock
+    scripts/build_paper.sh
+    scripts/generate_report.py
+    scripts/release_paper_artifacts.sh
+    scripts/run_paper_experiments.py
+    paper/main.tex
+    paper/anonymous.tex
+    paper/checklist.tex
+    paper/references.bib
+    paper/neurips_2026.sty
+  )
+  if ! git -C "${repo_root}" diff --quiet \
+    "${reviewed_source_commit}" "${current_commit}" -- \
+    "${release_source_paths[@]}"; then
+    printf '%s\n' \
+      "QUANTCORTEX_GENERATED_AT is required for changed release source" >&2
+    exit 1
+  fi
+  source_commit="${reviewed_source_commit}"
+  generated_at="${reviewed_generated_at}"
+fi
 source_date_epoch="${SOURCE_DATE_EPOCH:-$(git -C "${repo_root}" show -s --format=%ct "${source_commit}")}"
 temporary_root="$(mktemp -d "${TMPDIR:-/tmp}/quantcortex-paper-release.XXXXXX")"
 source_worktree="${temporary_root}/source"
@@ -44,10 +129,26 @@ trap cleanup EXIT
 git -C "${repo_root}" worktree add --detach "${source_worktree}" "${source_commit}" \
   >/dev/null
 
-permission="Owner-supplied local input; publication of derived aggregate "
-permission+="results authorized by the repository owner; provider terms not "
-permission+="independently verified"
-adjustment="yfinance adjusted close with auto_adjust=False"
+provider="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["source"]["provider"])' \
+    "${reviewed_manifest}"
+)"
+permission="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["source"]["permission_basis"])' \
+    "${reviewed_manifest}"
+)"
+retrieved_at="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["source"]["retrieved_at"])' \
+    "${reviewed_manifest}"
+)"
+adjustment="$(
+  "${python_bin}" -c \
+    'import json, sys; print(json.load(open(sys.argv[1]))["source"]["adjustment_method"])' \
+    "${reviewed_manifest}"
+)"
 
 (
   cd "${source_worktree}"
@@ -58,9 +159,9 @@ adjustment="yfinance adjusted close with auto_adjust=False"
     --cash-proxy-symbol SHV \
     --output-dir "${generated_output}" \
     --bootstrap-replications 5000 \
-    --data-provider "Yahoo Finance via yfinance 1.4.1" \
+    --data-provider "${provider}" \
     --permission-basis "${permission}" \
-    --retrieved-at 2026-06-16 \
+    --retrieved-at "${retrieved_at}" \
     --adjustment-method "${adjustment}" \
     --generated-at "${generated_at}" \
     --require-clean-source
@@ -76,9 +177,9 @@ adjustment="yfinance adjusted close with auto_adjust=False"
     --imgdir "${performance_output}/img" \
     --report-out "${performance_output}/report.md" \
     --manifest-out "${performance_output}/img/performance_manifest.json" \
-    --data-provider "Yahoo Finance via yfinance 1.4.1" \
+    --data-provider "${provider}" \
     --permission-basis "${permission}" \
-    --retrieved-at 2026-06-16 \
+    --retrieved-at "${retrieved_at}" \
     --adjustment-method "${adjustment}" \
     --generated-at "${generated_at}" \
     --require-clean-source \
diff --git a/scripts/run_paper_experiments.py b/scripts/run_paper_experiments.py
index d9f9c89..9ef0870 100644
--- a/scripts/run_paper_experiments.py
+++ b/scripts/run_paper_experiments.py
@@ -25,6 +25,12 @@
 logging.getLogger("hmmlearn").setLevel(logging.ERROR)
 os.environ.setdefault("LOKY_MAX_CPU_COUNT", "1")
 
+from quantcortex.backtest.conformance import (
+    target_tape_from_payload,
+    target_tape_to_payload,
+    target_tape_to_weights,
+    weights_to_target_tape,
+)
 from quantcortex.backtest.costs.transaction_costs import TransactionCostModel
 from quantcortex.backtest.engines.event_driven import EventDrivenBacktest
 from quantcortex.backtest.engines.vectorized import BacktestResult, VectorizedBacktest
@@ -78,7 +84,7 @@
 PRIMARY_BOOTSTRAP_BLOCK_LENGTH = 21
 BOOTSTRAP_SEED = 42
 PAPER_MAX_FORWARD_FILL = 0
-PROVENANCE_SCHEMA_VERSION = 4
+PROVENANCE_SCHEMA_VERSION = 5
 EVALUATION_CONTRACT_SCHEMA_VERSION = 1
 TARGET_TAPE_SCHEMA_VERSION = 1
 TARGET_EXPOSURE_COMPARATOR_SYMBOL = "equal_initial_weight_basket"
@@ -90,6 +96,7 @@
     "net_excess_over_cash": "Net excess over cash",
 }
 SOURCE_TREE_FIXED_FILES = (
+    "scripts/release_paper_artifacts.sh",
     "scripts/run_paper_experiments.py",
     "schemas/canonical_target_tape.schema.json",
     "schemas/evaluation_contract.schema.json",
@@ -152,6 +159,32 @@ def _git_metadata(repo_root: Path) -> dict[str, str | bool]:
     }
 
 
+def _git_path_is_tracked(repo_root: Path, path: Path) -> bool:
+    """Return whether ``path`` is tracked in the source repository."""
+    try:
+        relative = path.resolve().relative_to(repo_root.resolve()).as_posix()
+    except ValueError:
+        return False
+    try:
+        result = subprocess.run(
+            ["git", "ls-files", "--error-unmatch", "--", relative],
+            cwd=repo_root,
+            check=False,
+            capture_output=True,
+            text=True,
+        )
+    except OSError as exc:
+        raise RuntimeError("could not determine whether paper input is tracked") from exc
+    if result.returncode == 0:
+        return True
+    if result.returncode == 1:
+        return False
+    detail = result.stderr.strip() or f"git exited with status {result.returncode}"
+    raise RuntimeError(
+        f"could not determine whether paper input is tracked: {detail}"
+    )
+
+
 def _sha256(path: Path) -> str:
     digest = hashlib.sha256()
     with path.open("rb") as handle:
@@ -1756,6 +1789,7 @@ def run_experiments(
         evaluation_end,
     )
     strategy_weights: dict[str, pd.DataFrame] = {}
+    target_tape_metadata: dict[str, dict[str, object]] = {}
     ablation_rows: list[dict[str, float | str]] = []
     variant_series: dict[str, dict[str, pd.Series]] = {}
     gross_active_by_variant: dict[str, pd.Series] = {}
@@ -1776,8 +1810,28 @@ def run_experiments(
         )
 
     for variant, strategy in strategies.items():
-        weights = strategy.generate_weights(prices, weekly)
+        raw_weights = strategy.generate_weights(prices, weekly)
+        target_tape = weights_to_target_tape(raw_weights, max_gross=1.0)
+        target_payload = target_tape_to_payload(
+            target_tape,
+            max_gross=1.0,
+            expected_symbols=list(prices.columns),
+        )
+        validated_tape = target_tape_from_payload(target_payload)
+        weights = target_tape_to_weights(
+            validated_tape,
+            max_gross=1.0,
+            expected_symbols=list(prices.columns),
+        )
         strategy_weights[variant] = weights
+        target_tape_metadata[variant] = {
+            "schema_version": TARGET_TAPE_SCHEMA_VERSION,
+            "decision_count": int(weights.shape[0]),
+            "record_count": int(len(target_tape)),
+            "symbols": list(target_payload["symbols"]),
+            "max_gross": 1.0,
+            "canonical_payload_sha256": _json_sha256(target_payload),
+        }
         result = _engine_result(
             weights,
             prices,
@@ -2088,6 +2142,7 @@ def protocol_row(
         "evaluation_index": evaluation_index,
         "warmup_sessions": available_warmup_sessions,
         "required_warmup_sessions": required_warmup_sessions,
+        "target_tape_metadata": target_tape_metadata,
         "accounting": pd.DataFrame(accounting_rows),
         "ablation": ablation,
         "ablation_uncertainty": ablation_uncertainty,
@@ -2160,12 +2215,16 @@ def main(argv: list[str]) -> int:
         raise ValueError(f"dependency lock is missing: {dependency_lock_path}")
 
     source = args.prices_csv.expanduser().resolve()
+    raw_input_committed = _git_path_is_tracked(repo_root, source)
+    if raw_input_committed:
+        raise RuntimeError("paper price input must not be tracked by Git")
     input_digest = sha256_file(source)
     symbols = UNIVERSE + [args.cash_proxy_symbol]
     prices_with_cash = load_price_matrix(
         source,
         symbols=symbols,
         max_ffill=PAPER_MAX_FORWARD_FILL,
+        require_complete=True,
     )
     prices = prices_with_cash.loc[:, UNIVERSE]
     cash_returns = prices_with_cash[args.cash_proxy_symbol].pct_change(
@@ -2233,6 +2292,10 @@ def main(argv: list[str]) -> int:
             path=results_dir / "generated_values.tex",
         ),
         _write_json(contract, results_dir / "evaluation_contract.json"),
+        _write_json(
+            experiment["target_tape_metadata"],
+            results_dir / "target_tape_hashes.json",
+        ),
     ]
     uncertainty_path = results_dir / "uncertainty.json"
     artifacts.append(_write_json(experiment["uncertainty"], uncertainty_path))
@@ -2286,7 +2349,7 @@ def main(argv: list[str]) -> int:
         "source": {
             "file_name": source.name,
             "input_sha256": input_digest,
-            "raw_input_committed": False,
+            "raw_input_committed": raw_input_committed,
             "symbols": symbols,
             "cash_proxy": args.cash_proxy_symbol,
             "provider": args.data_provider,
@@ -2414,6 +2477,10 @@ def main(argv: list[str]) -> int:
             "schema_version": EVALUATION_CONTRACT_SCHEMA_VERSION,
             "canonical_sha256": _json_sha256(contract),
         },
+        "decision_streams": {
+            "path": "results/target_tape_hashes.json",
+            "variants": experiment["target_tape_metadata"],
+        },
         "artifacts": {
             str(path.relative_to(output_dir)): _sha256(path)
             for path in sorted(artifacts)
diff --git a/tests/test_backtest_conformance.py b/tests/test_backtest_conformance.py
index f55b0a3..68f1374 100644
--- a/tests/test_backtest_conformance.py
+++ b/tests/test_backtest_conformance.py
@@ -1,10 +1,13 @@
 from __future__ import annotations
 
+import copy
 import json
 from pathlib import Path
 
+import numpy as np
 import pandas as pd
 import pytest
+from jsonschema import Draft202012Validator, FormatChecker
 
 from quantcortex.backtest.conformance import (
     TARGET_TAPE_COLUMNS,
@@ -26,6 +29,89 @@ def _fixture_tape() -> pd.DataFrame:
     return pd.read_csv(FIXTURE_ROOT / "target_tape.csv")
 
 
+def _target_tape_schema_validator() -> Draft202012Validator:
+    schema = json.loads(
+        (REPO_ROOT / "schemas" / "canonical_target_tape.schema.json").read_text()
+    )
+    return Draft202012Validator(schema, format_checker=FormatChecker())
+
+
+def _wire_case_payload(case: str) -> dict[str, object]:
+    payload = copy.deepcopy(
+        target_tape_to_payload(_fixture_tape(), expected_symbols=["A", "B"])
+    )
+    records = payload["records"]
+    assert isinstance(records, list)
+
+    if case == "valid":
+        return payload
+    if case == "integral_float_schema_version":
+        payload["schema_version"] = 1.0
+    elif case == "boolean_schema_version":
+        payload["schema_version"] = True
+    elif case == "nonintegral_schema_version":
+        payload["schema_version"] = 1.5
+    elif case == "unsupported_schema_version":
+        payload["schema_version"] = 2
+    elif case == "string_max_gross":
+        payload["max_gross"] = "1.0"
+    elif case == "negative_target_weight":
+        records[0]["target_weight"] = -5e-13
+    elif case == "oversized_binary64_number":
+        oversized = 10**1000
+        payload["max_gross"] = oversized
+        records[0]["target_weight"] = oversized
+    elif case == "maximum_binary64_number":
+        maximum = float.fromhex("0x1.fffffffffffffp+1023")
+        payload["max_gross"] = maximum
+        records[0]["target_weight"] = maximum
+    elif case in {
+        "date_only_timestamp",
+        "lowercase_timestamp",
+        "year_zero_timestamp",
+        "invalid_calendar_date",
+        "fractional_timestamp",
+        "maximum_offset_timestamp",
+        "local_hour_24",
+        "local_minute_60",
+        "local_second_60",
+        "offset_hour_24",
+        "offset_minute_60",
+    }:
+        replacements = {
+            "date_only_timestamp": "2024-01-01",
+            "lowercase_timestamp": "2024-01-01t00:00:00z",
+            "year_zero_timestamp": "0000-01-01T00:00:00Z",
+            "invalid_calendar_date": "2023-02-29T00:00:00Z",
+            "fractional_timestamp": "2024-01-01T00:00:00.123456789Z",
+            "maximum_offset_timestamp": "2024-01-01T00:00:00+23:59",
+            "local_hour_24": "2024-01-01T24:00:00Z",
+            "local_minute_60": "2024-01-01T00:60:00Z",
+            "local_second_60": "2024-01-01T00:00:60Z",
+            "offset_hour_24": "2024-01-01T00:00:00+24:00",
+            "offset_minute_60": "2024-01-01T00:00:00+00:60",
+        }
+        original = records[0]["decision_timestamp"]
+        for record in records:
+            if record["decision_timestamp"] == original:
+                record["decision_timestamp"] = replacements[case]
+    elif case in {"blank_symbol", "outer_whitespace_symbol"}:
+        replacement = " " if case == "blank_symbol" else " A"
+        symbols = payload["symbols"]
+        assert isinstance(symbols, list)
+        symbols[0] = replacement
+        for record in records:
+            if record["symbol"] == "A":
+                record["symbol"] = replacement
+    elif case == "unknown_payload_field":
+        payload["unexpected"] = True
+    elif case == "unknown_record_field":
+        records[0]["unexpected"] = True
+    else:  # pragma: no cover - protects the test table itself
+        raise AssertionError(f"unknown wire test case: {case}")
+    return payload
+
+
 def test_canonical_target_tape_round_trips():
     tape = validate_target_tape(_fixture_tape(), expected_symbols=["A", "B"])
     weights = target_tape_to_weights(tape, expected_symbols=["A", "B"])
@@ -51,6 +137,73 @@ def test_canonical_target_tape_json_payload_round_trips():
     pd.testing.assert_frame_equal(restored, tape)
 
 
+@pytest.mark.parametrize(
+    ("case", "accepted"),
+    [
+        ("valid", True),
+        ("integral_float_schema_version", True),
+        ("boolean_schema_version", False),
+        ("nonintegral_schema_version", False),
+        ("unsupported_schema_version", False),
+        ("string_max_gross", False),
+        ("negative_target_weight", False),
+        ("oversized_binary64_number", False),
+        ("maximum_binary64_number", True),
+        ("date_only_timestamp", False),
+        ("lowercase_timestamp", False),
+        ("year_zero_timestamp", False),
+        ("invalid_calendar_date", False),
+        ("fractional_timestamp", True),
+        ("maximum_offset_timestamp", True),
+        ("local_hour_24", False),
+        ("local_minute_60", False),
+        ("local_second_60", False),
+        ("offset_hour_24", False),
+        ("offset_minute_60", False),
+        ("blank_symbol", False),
+        ("outer_whitespace_symbol", False),
+        ("unknown_payload_field", False),
+        ("unknown_record_field", False),
+    ],
+)
+def test_target_tape_schema_and_runtime_align_on_wire_constraints(case, accepted):
+    """Keep primitive wire validation aligned without duplicating semantics."""
+    payload = _wire_case_payload(case)
+    schema_accepts = not list(_target_tape_schema_validator().iter_errors(payload))
+    try:
+        target_tape_from_payload(payload)
+    except (TypeError, ValueError):
+        runtime_accepts = False
+    else:
+        runtime_accepts = True
+
+    assert schema_accepts is accepted
+    assert runtime_accepts is accepted
+
+
+def test_canonical_target_tape_payload_survives_strict_json_round_trip():
+    tape = validate_target_tape(_fixture_tape(), expected_symbols=["A", "B"])
+    payload = target_tape_to_payload(tape, expected_symbols=["A", "B"])
+    encoded = json.dumps(payload, allow_nan=False, sort_keys=True)
+    decoded = json.loads(encoded)
+
+    _target_tape_schema_validator().validate(decoded)
+    restored = target_tape_from_payload(decoded)
+
+    pd.testing.assert_frame_equal(restored, tape)
+
+
+@pytest.mark.parametrize("value", [float("nan"), float("inf"), float("-inf")])
+def test_canonical_target_tape_payload_rejects_nonfinite_json_numbers(value):
+    payload = target_tape_to_payload(_fixture_tape(), expected_symbols=["A", "B"])
+    payload["records"][0]["target_weight"] = value
+
+    with pytest.raises(ValueError, match="must be finite binary64"):
+        target_tape_from_payload(payload)
+    with pytest.raises(ValueError, match="Out of range float values"):
+        json.dumps(payload, allow_nan=False)
+
+
 @pytest.mark.parametrize(
     ("mutator", "message"),
     [
@@ -88,6 +241,81 @@ def test_canonical_target_tape_payload_rejects_unknown_fields():
         target_tape_from_payload(payload)
 
 
+def test_target_tape_to_weights_preserves_declared_symbol_order():
+    weights = target_tape_to_weights(
+        _fixture_tape(),
+        expected_symbols=["B", "A"],
+    )
+
+    assert list(weights.columns) == ["B", "A"]
+
+
+def test_target_tape_long_only_tolerance_preserves_positive_weights_only():
+    tape = _fixture_tape()
+    tape.loc[0, "target_weight"] = 5e-16
+    tape.loc[1, "target_weight"] = -5e-13
+
+    normalized = validate_target_tape(tape, expected_symbols=["A", "B"])
+
+    assert normalized.loc[0, "target_weight"] == 5e-16
+    assert normalized.loc[1, "target_weight"] == 0.0
+    payload = target_tape_to_payload(normalized, expected_symbols=["A", "B"])
+    roundtrip = target_tape_to_weights(
+        target_tape_from_payload(payload),
+        expected_symbols=["A", "B"],
+    )
+    assert roundtrip.iloc[0].to_dict() == {"A": 5e-16, "B": 0.0}
+
+    tape.loc[1, "target_weight"] = -2e-12
+    with pytest.raises(ValueError, match="long-only"):
+        validate_target_tape(tape, expected_symbols=["A", "B"])
+
+
+@pytest.mark.parametrize(
+    ("path", "value", "error", "message"),
+    [
+        (("schema_version",), True, TypeError, "schema_version must be a JSON number"),
+        (("max_gross",), "1.0", TypeError, "max_gross must be a JSON number"),
+        (
+            ("records", 0, "decision_timestamp"),
+            "2024-01-01",
+            ValueError,
+            "must be an RFC 3339 date-time",
+        ),
+        (
+            ("records", 0, "decision_timestamp"),
+            1_704_067_200,
+            TypeError,
+            "must be an RFC 3339 string",
+        ),
+        (
+            ("records", 0, "target_weight"),
+            "1.0",
+            TypeError,
+            "target_weight must be a JSON number",
+        ),
+        (
+            ("records", 0, "target_weight"),
+            np.float64(1.0),
+            TypeError,
+            "target_weight must be a JSON number",
+        ),
+    ],
+)
+def test_canonical_target_tape_payload_enforces_published_json_types(
+    path, value, error, message
+):
+    payload = target_tape_to_payload(_fixture_tape(), expected_symbols=["A", "B"])
+    mutated = copy.deepcopy(payload)
+    cursor = mutated
+    for key in path[:-1]:
+        cursor = cursor[key]
+    cursor[path[-1]] = value
+
+    with pytest.raises(error, match=message):
+        target_tape_from_payload(mutated)
+
+
 def test_conformance_fixture_has_hand_computable_next_bar_returns():
     weights = target_tape_to_weights(
         _fixture_tape(),
@@ -109,9 +337,7 @@ def test_conformance_fixture_has_hand_computable_next_bar_returns():
 
 
 def test_committed_contract_schemas_are_versioned_and_specific():
-    target_schema = json.loads(
-        (REPO_ROOT / "schemas" / "canonical_target_tape.schema.json").read_text()
-    )
+    target_schema = _target_tape_schema_validator().schema
     contract_schema = json.loads(
         (REPO_ROOT / "schemas" / "evaluation_contract.schema.json").read_text()
     )
@@ -136,3 +362,13 @@ def test_committed_contract_schemas_are_versioned_and_specific():
     assert contract_schema["properties"]["target_tape"][
         "additionalProperties"
     ] is False
+
+    _target_tape_schema_validator().validate(
+        target_tape_to_payload(_fixture_tape(), expected_symbols=["A", "B"])
+    )
+    contract_validator = Draft202012Validator(contract_schema)
+    contract_validator.validate(
+        json.loads(
+            (REPO_ROOT / "paper" / "results" / "evaluation_contract.json").read_text()
+        )
+    )
diff --git a/tests/test_local_csv.py b/tests/test_local_csv.py
index 920f2cb..c060709 100644
--- a/tests/test_local_csv.py
+++ b/tests/test_local_csv.py
@@ -70,6 +70,31 @@ def test_load_price_matrix_rejects_missing_symbol(tmp_path):
         load_price_matrix(path, max_ffill=True)
     with pytest.raises(LocalDataError, match="valid timestamp"):
         load_price_matrix(path, start="not-a-date")
+    with pytest.raises(LocalDataError, match="must be a boolean"):
+        load_price_matrix(path, require_complete=1)
+
+
+def test_load_price_matrix_can_fail_closed_on_incomplete_rows(tmp_path):
+    path = tmp_path / "prices.csv"
+    pd.DataFrame(
+        {
+            "date": ["2024-01-01", "2024-01-02", "2024-01-03"],
+            "AAA": [10.0, None, 12.0],
+            "BBB": [20.0, 21.0, 22.0],
+        }
+    ).to_csv(path, index=False)
+
+    permissive = load_price_matrix(path, max_ffill=0)
+    assert permissive.index.tolist() == [
+        pd.Timestamp("2024-01-01"),
+        pd.Timestamp("2024-01-03"),
+    ]
+
+    with pytest.raises(
+        LocalDataError,
+        match=r"missing required observations.*2024-01-02:AAA",
+    ):
+        load_price_matrix(path, max_ffill=0, require_complete=True)
 
 
 def test_load_price_matrix_rejects_duplicate_csv_headers(tmp_path):
diff --git a/tests/test_paper_artifacts.py b/tests/test_paper_artifacts.py
index 7f186d7..ba0523b 100644
--- a/tests/test_paper_artifacts.py
+++ b/tests/test_paper_artifacts.py
@@ -22,7 +22,7 @@ def test_paper_artifacts_match_manifest_and_generator():
     manifest = json.loads(
         (PAPER_ROOT / "results" / "manifest.json").read_text(encoding="utf-8")
     )
-    assert manifest["schema_version"] == 4
+    assert manifest["schema_version"] == 5
 
     source = manifest["source"]
     assert source["raw_input_committed"] is False
@@ -51,6 +51,7 @@ def test_paper_artifacts_match_manifest_and_generator():
             source_commit,
             "--",
             "quantcortex",
+            "scripts/release_paper_artifacts.sh",
             "scripts/run_paper_experiments.py",
             "schemas",
             "pyproject.toml",
@@ -64,6 +65,7 @@ def test_paper_artifacts_match_manifest_and_generator():
     assert dependency_lock["sha256"] == _sha256(REPO_ROOT / "poetry.lock")
     assert generator["threadpools"]
     source_tree = generator["source_tree"]
+    assert "scripts/release_paper_artifacts.sh" in source_tree["files"]
     assert source_tree == source_tree_manifest(
         REPO_ROOT,
         list(source_tree["files"]),
@@ -157,6 +159,7 @@ def test_paper_artifacts_match_manifest_and_generator():
     assert "results/ablation_uncertainty.csv" in artifacts
     assert "results/comparator_diagnostics.csv" in artifacts
     assert "results/evaluation_contract.json" in artifacts
+    assert "results/target_tape_hashes.json" in artifacts
     assert "results/return_decomposition.csv" in artifacts
     assert "results/sharpe_uncertainty.csv" in artifacts
     assert "results/protocol_switches.csv" in artifacts
@@ -167,6 +170,18 @@ def test_paper_artifacts_match_manifest_and_generator():
         assert artifact.is_file(), relative_path
         assert _sha256(artifact) == expected_digest, relative_path
 
+    target_tape_path = PAPER_ROOT / manifest["decision_streams"]["path"]
+    target_tapes = json.loads(target_tape_path.read_text(encoding="utf-8"))
+    assert target_tapes == manifest["decision_streams"]["variants"]
+    assert set(target_tapes) == {"full", "no_regime", "no_vol_scaler", "signal_only"}
+    for metadata in target_tapes.values():
+        assert metadata["symbols"] == sorted(metadata["symbols"])
+        assert metadata["record_count"] == (
+            metadata["decision_count"] * len(metadata["symbols"])
+        )
+        assert len(metadata["canonical_payload_sha256"]) == 64
+        int(metadata["canonical_payload_sha256"], 16)
+
     with (PAPER_ROOT / "results" / "protocol_switches.csv").open(
         encoding="utf-8",
         newline="",
@@ -231,6 +246,9 @@ def test_paper_artifacts_match_manifest_and_generator():
 
 def test_paper_source_and_reviewed_pdf_are_published():
     main = (PAPER_ROOT / "main.tex").read_text(encoding="utf-8")
+    release_script = (REPO_ROOT / "scripts" / "release_paper_artifacts.sh").read_text(
+        encoding="utf-8"
+    )
 
     assert "\\usepackage[preprint]{neurips_2026}" in main
     assert "\\usepackage{orcidlink}" in main
@@ -243,6 +261,17 @@ def test_paper_source_and_reviewed_pdf_are_published():
     assert "\\PaperInputDigest" in main
     assert "{bootstrap_robustness.pdf}" in main
     assert "target-exposure comparator" in main
+    assert "actual_input_digest" in release_script
+    assert "expected_input_digest" in release_script
+    assert "ls-files --error-unmatch" in release_script
+    assert "reviewed_generated_at" in release_script
+    assert "reviewed_source_commit" in release_script
+    assert "release_source_paths" in release_script
+    assert "scripts/release_paper_artifacts.sh" in release_script
+    assert "QUANTCORTEX_GENERATED_AT is required for changed release source" in (
+        release_script
+    )
+    assert '--data-provider "${provider}"' in release_script
 
     anonymous_source = (PAPER_ROOT / "anonymous.tex").read_text(encoding="ascii")
     assert "\\def\\quantcortexanonymous{1}" in anonymous_source
diff --git a/tests/test_paper_experiments.py b/tests/test_paper_experiments.py
index 9fd88ee..0d4b7df 100644
--- a/tests/test_paper_experiments.py
+++ b/tests/test_paper_experiments.py
@@ -4,6 +4,7 @@
 import json
 import subprocess
 from pathlib import Path
+from types import SimpleNamespace
 
 import numpy as np
 import pandas as pd
@@ -20,6 +21,7 @@
     _cagr,
     _costed_target_exposure_comparator,
     _git_metadata,
+    _git_path_is_tracked,
     _max_drawdown,
     _save_figures,
     _tex_number,
@@ -385,6 +387,33 @@ def test_git_metadata_captures_cleanliness_before_writes(tmp_path):
     assert dirty["worktree_clean_at_start"] is False
 
 
+def test_git_path_tracking_is_measured_from_the_source_repository(tmp_path):
+    subprocess.run(["git", "init", "-q"], cwd=tmp_path, check=True)
+    tracked = tmp_path / "tracked.csv"
+    untracked = tmp_path / "untracked.csv"
+    outside = tmp_path.parent / f"{tmp_path.name}-outside.csv"
+    tracked.write_text("date,A\n2024-01-01,1\n", encoding="ascii")
+    untracked.write_text("date,A\n2024-01-01,1\n", encoding="ascii")
+    outside.write_text("date,A\n2024-01-01,1\n", encoding="ascii")
+    subprocess.run(["git", "add", "tracked.csv"], cwd=tmp_path, check=True)
+
+    assert _git_path_is_tracked(tmp_path, tracked) is True
+    assert _git_path_is_tracked(tmp_path, untracked) is False
+    assert _git_path_is_tracked(tmp_path, outside) is False
+
+
+def test_git_path_tracking_fails_closed_on_git_errors(monkeypatch, tmp_path):
+    candidate = tmp_path / "prices.csv"
+    candidate.write_text("date,A\n2024-01-01,1\n", encoding="ascii")
+    monkeypatch.setattr(
+        "scripts.run_paper_experiments.subprocess.run",
+        lambda *args, **kwargs: SimpleNamespace(returncode=128, stderr="fatal"),
+    )
+
+    with pytest.raises(RuntimeError, match="could not determine"):
+        _git_path_is_tracked(tmp_path, candidate)
+
+
 def test_paper_cli_requires_clean_source_before_reading_input(monkeypatch, tmp_path):
     monkeypatch.setattr(
         "scripts.run_paper_experiments._git_metadata",
@@ -412,6 +441,36 @@ def test_paper_cli_requires_clean_source_before_reading_input(monkeypatch, tmp_p
         )
 
 
+def test_paper_cli_rejects_a_git_tracked_price_input(monkeypatch):
+    monkeypatch.setattr(
+        "scripts.run_paper_experiments._git_metadata",
+        lambda _repo_root: {
+            "source_commit": "0" * 40,
+            "worktree_clean_at_start": True,
+        },
+    )
+    tracked_fixture = (
+        Path(__file__).resolve().parent / "fixtures" / "conformance" / "prices.csv"
+    )
+
+    with pytest.raises(RuntimeError, match="must not be tracked by Git"):
+        paper_main(
+            [
+                "run_paper_experiments.py",
+                "--prices-csv",
+                str(tracked_fixture),
+                "--data-provider",
+                "test provider",
+                "--permission-basis",
+                "test permission",
+                "--retrieved-at",
+                "2026-06-16",
+                "--adjustment-method",
+                "test adjustment",
+            ]
+        )
+
+
 def test_paper_cli_writes_a_complete_test_only_artifact_set(monkeypatch, tmp_path):
     warmup = pd.bdate_range(end="2017-12-29", periods=300)
     first_period = pd.bdate_range("2018-01-02", periods=40)
@@ -442,6 +501,16 @@ def test_paper_cli_writes_a_complete_test_only_artifact_set(monkeypatch, tmp_pat
         "scripts.run_paper_experiments._threadpool_environment",
         lambda: [{"user_api": "test", "num_threads": 1}],
     )
+    from scripts import run_paper_experiments as paper_experiments
+
+    parsed_payloads = []
+    original_parser = paper_experiments.target_tape_from_payload
+
+    def recording_parser(payload):
+        parsed_payloads.append(payload)
+        return original_parser(payload)
+
+    monkeypatch.setattr(paper_experiments, "target_tape_from_payload", recording_parser)
 
     assert (
         paper_main(
@@ -476,7 +545,7 @@ def test_paper_cli_writes_a_complete_test_only_artifact_set(monkeypatch, tmp_pat
     manifest = json.loads(
         (output_dir / "results" / "manifest.json").read_text(encoding="utf-8")
     )
-    assert manifest["schema_version"] == 4
+    assert manifest["schema_version"] == 5
     assert manifest["generated_at"] == "2026-06-16T00:00:00Z"
     assert manifest["generator"]["git"] == {
         "source_commit": "1" * 40,
@@ -485,11 +554,24 @@ def test_paper_cli_writes_a_complete_test_only_artifact_set(monkeypatch, tmp_pat
     assert manifest["source"]["raw_input_committed"] is False
     assert manifest["source"]["provider"] == "deterministic synthetic test fixture"
     assert "results/evaluation_contract.json" in manifest["artifacts"]
+    assert "results/target_tape_hashes.json" in manifest["artifacts"]
     assert "results/sharpe_uncertainty.csv" in manifest["artifacts"]
     assert "figures/accounting_summary.pdf" in manifest["artifacts"]
     for relative_path in manifest["artifacts"]:
         assert (output_dir / relative_path).is_file(), relative_path
 
+    target_tapes = manifest["decision_streams"]["variants"]
+    assert set(target_tapes) == {"full", "no_regime", "no_vol_scaler", "signal_only"}
+    assert all(metadata["schema_version"] == 1 for metadata in target_tapes.values())
+    assert all(metadata["decision_count"] > 0 for metadata in target_tapes.values())
+    assert all(metadata["symbols"] == sorted(metadata["symbols"]) for metadata in target_tapes.values())
+    assert all(
+        metadata["record_count"]
+        == metadata["decision_count"] * len(metadata["symbols"])
+        for metadata in target_tapes.values()
+    )
+    assert len(parsed_payloads) == 4
+
 
 def test_evaluation_contract_separates_attribution_and_tradable_comparator():
     contract = evaluation_contract("SHV")
diff --git a/tests/test_repository_data_policy.py b/tests/test_repository_data_policy.py
index 66345ba..017039a 100644
--- a/tests/test_repository_data_policy.py
+++ b/tests/test_repository_data_policy.py
@@ -4,6 +4,7 @@
 import hashlib
 import json
 import re
+import subprocess
 from pathlib import Path
 
 REPO_ROOT = Path(__file__).resolve().parent.parent
@@ -42,6 +43,28 @@ def test_redistributed_market_data_is_absent():
     present = [path for path in FORBIDDEN_MARKET_DATA if (REPO_ROOT / path).exists()]
     assert not present, f"redistributed market-data snapshots reappeared: {present}"
 
+    tracked = subprocess.run(
+        ["git", "ls-files"],
+        cwd=REPO_ROOT,
+        check=True,
+        capture_output=True,
+        text=True,
+    ).stdout.splitlines()
+    assert [path for path in tracked if path.startswith("local_data/")] == [
+        "local_data/README.md"
+    ]
+
+    raw_columns = {"date", "QQQ", "VGT", "GLD", "TLT", "SPY", "VIG", "SHV"}
+    leaked = []
+    for relative_path in tracked:
+        if not relative_path.endswith(".csv"):
+            continue
+        with (REPO_ROOT / relative_path).open(encoding="utf-8", newline="") as handle:
+            header = set(next(csv.reader(handle), []))
+        if raw_columns <= header:
+            leaked.append(relative_path)
+    assert not leaked, f"tracked raw paper price matrices detected: {leaked}"
+
 
 def test_published_performance_charts_match_manifest_and_readme():
     image_dir = REPO_ROOT / "docs" / "img"