Green Guides — Content Validation Report¶
Date: 2026-03-31
Scope: All content in inputs/ verified against Sources/ prior to publication
Validator: Athena Agent (Claude Opus 4.6)
Basis: .agents/ENTER/07-data-validity-and-qa.md
Executive Summary¶
| Module | Claims Verified | PASS | WEAK | FAIL | Coverage Gaps |
|---|---|---|---|---|---|
| regulatory-explorer | 21 | 20 | 1 | 0 | Several uncovered assertions |
| engineering-constraints | 18 | 18 | 0 | 0 | Minor |
| bmp-library (Tier 1 — 6 core BMPs) | 21 | 21 | 0 | 0 | None critical |
| bmp-library (Tier 2 — 5 BMPs) | 15 | 12 | 3 | 0 | Several uncovered |
| bmp-library (Tier 3 — 18 manifests) | ~36 | 30 | 5 | 1 | Notable gaps in ch10/ch11 index |
| design-workflow | 20 | 20 | 0 | 0 | None |
| glossary | 6 | 6 | 0 | 0 | None |
| TOTAL | 137 | 127 | 9 | 1 |
Overall readiness: CONDITIONAL. The content is fundamentally accurate and source-backed. Two FAIL-level issues require correction before publication. Nine WEAK findings should be reviewed. No claim was found to publish demonstrably false information to users, but dry-well.md contains one unsourced technical value that must be removed or verified.
Phase 0: Structural Integrity¶
Ran validate_input_claim_manifests.py and validate_citations.py.
- 48 manifests checked. 0 structural issues.
- All
source_markdown,source_pdf, andsource_manifestpaths resolve. - All manifests conform to
authored_claim_manifest.v1schema.
Result: CLEAN.
Phase 2A: Regulatory-Explorer (21 claims)¶
| Claim ID | Page | Verdict | Notes |
|---|---|---|---|
| reg-major-001 | major-development-definition | PASS | 1-acre/0.25-acre thresholds confirmed verbatim |
| reg-major-002 | major-development-definition | WEAK | Motor vehicle surface trigger dates to March 2021, not 2026. Grouped with reconstruction trigger as "2026 definition package" is misleading. Quote field also doesn't match this claim's subject. |
| reg-wq-001 | water-quality-design-storm | PASS | "1.25 inches of rainfall in two hours" confirmed verbatim at 7:8-5.5(d) |
| reg-wq-002 | water-quality-design-storm | PASS | 80% general standard / 50% narrow public-transportation exception confirmed |
| reg-wq-003 | water-quality-design-storm | PASS | 7:8-5.6(d) absent from 2023 source; WQDS value unchanged between eras |
| reg-gi-001 | green-infrastructure-requirement | PASS | GI definition confirmed verbatim from 7:8-1.2 |
| reg-gi-002 | green-infrastructure-requirement | PASS | 7:8-5.3(b) Table 5-1 GI directive confirmed |
| reg-gi-003 | green-infrastructure-requirement | PASS | 7:8-5.3(c)-(d) Table 5-2/5-3 structure and variance requirement confirmed |
| reg-gi-004 | green-infrastructure-requirement | PASS | "Non-GI BMP" and "volumetric reduction" absent from 7:8-1.2 definitions section |
| reg-gi-005 | green-infrastructure-requirement | PASS | 7:8-5.9 is "Sources for technical guidance," not volumetric reduction |
| reg-vr-001 | volumetric-reduction-credit | PASS | 7:8-5.6(d) language confirmed verbatim |
| reg-vr-002 | volumetric-reduction-credit | PASS | WQDS retention via Table 5-1/5-2 confirmed in NJAC and BMP Ch.14 |
| reg-vr-003 | volumetric-reduction-credit | PASS | HUC-14 constraint, concurrent-construction requirement, offsite options all confirmed |
| reg-quantity-001 | stormwater-quantity-control | PASS | 2/10/100-yr storms at 50%/75%/80% confirmed verbatim at 7:8-5.6(b)3 |
| reg-quantity-002 | stormwater-quantity-control | PASS | 7:8-5.6(d) confirmed as 2026-only addition; absent from 2023 |
| reg-quantity-003 | stormwater-quantity-control | PASS | NOAA Atlas 14 + Tables 5-5/5-6 confirmed at 7:8-5.7 |
| reg-quantity-004 | stormwater-quantity-control | PASS | BMP Ch.5 confirms three quantity-compliance options |
| reg-recharge-001 | groundwater-recharge-rules | PASS | 100% pre-construction recharge maintenance confirmed at 7:8-5.4(b)1i |
| reg-recharge-002 | groundwater-recharge-rules | PASS | Urban redevelopment exemption and high-pollutant-loading prohibition confirmed |
| reg-recharge-003 | groundwater-recharge-rules | PASS | Wooded/good-condition presumption confirmed (minor precision note: source says "good cover" for woods specifically, not "good hydrologic condition") |
| reg-recharge-004 | groundwater-recharge-rules | PASS | BMP Ch.6 recharge workflow and multi-BMP guidance confirmed |
Uncovered assertions in .md files (not blocking, but coverage gaps):
- major-development-definition.md: reconstruction-is-2026 framing, common-plan-of-development provision not claimed
- water-quality-design-storm.md: cross-era stability of WQDS not explicitly claimed; volumetric reduction as "retention-focused" not claimed
- stormwater-quantity-control.md: "storm set unchanged 2023→2026" assertion not claimed
- groundwater-recharge-rules.md: NJGRS tool, coordination with SHWT/mounding/pollutant-loading not claimed
Phase 2B: Engineering-Constraints (18 claims)¶
All 18 claims PASS.
| Claim ID | Verdict | Key Evidence |
|---|---|---|
| eng-soil-001 | PASS | ASTM D3385 confirmed in Ch.12; D5126 absent from source |
| eng-soil-002 | PASS | K_design = K_field/2.0 confirmed in Ch.13 (formula and worked examples) |
| eng-soil-003 | PASS | "minimum depth of eight feet below the lowest elevation of the basin bottom" confirmed verbatim Ch.12 p.18 |
| eng-soil-004 | PASS | 1 pit per soil mapping unit, 2 borings per GI BMP >500 sf confirmed Ch.12 §3 Item 3 |
| eng-soil-005 | PASS | Post-construction non-test-hole methods confirmed verbatim Ch.12 p.29 |
| eng-shwt-001 | PASS | Highest redox level = SHWT unless higher observed water table controls, confirmed Ch.12 p.13 |
| eng-shwt-002 | PASS | Direct-observation fallback for no-redox sites confirmed Ch.12 pp.13-14 |
| eng-shwt-003 | PASS | 2-ft separation standard confirmed in Ch.9.6 p.11-13 (pervious paving); Ch.13 SHWT-at-zero confirmed p.3 |
| eng-mound-001 | PASS | 7:8-5.3(h) confirmed as correct hydraulic-impact trigger (Ch.13 intro quotes rule verbatim) |
| eng-mound-002 | PASS | No numeric triggers (2.99 in/hr, 5,000 cf, SHWT within 4 ft) in 2026 NJAC 7:8-5.3(h) |
| eng-mound-003 | PASS | Hantush Spreadsheet confirmed as the Ch.13 analysis tool |
| eng-mound-004 | PASS | "SHWT at elevation zero" confirmed verbatim in Ch.13 p.3 |
| eng-mound-005 | PASS | Design permeability (K_field/2) and 72-hr max duration confirmed in Ch.13 pp.4-5 |
| eng-da-001 | PASS | 2.5-acre small-scale bioretention limit confirmed in NJAC 7:8-5.3(b) Table 5-1 |
| eng-da-002 | PASS | 3:1 inflow-area-to-surface-area ratio confirmed Ch.9.6 p.3 |
| eng-infil-001 | PASS | Infiltration BMP category examples confirmed in Ch.12 p.16 |
| eng-infil-002 | PASS | Framework-based feasibility (not numeric shortcuts) confirmed Ch.12 purpose section |
| eng-infil-003 | PASS | High-pollutant-loading recharge prohibition confirmed at N.J.A.C. 7:8-5.4(b)3 |
Notable finding: BMP Manual Ch.13 (p.1) internally quotes the hydraulic-impact trigger as "N.J.A.C. 7:8-5.2(h)" — this appears to be an error in the manual itself; the NJAC source places the requirement at 7:8-5.3(h). The claims correctly cite 7:8-5.3(h).
Phase 2C: BMP Library¶
Tier 1 — Core 6 BMPs (21 claims, all PASS)¶
| BMP | Claims | Result |
|---|---|---|
| bioretention | bmp-bio-001 through 005 | All PASS. Media composition (85-95% sand) and design table values confirmed. E-12/E-13 crosswalk errors have been corrected. |
| pervious pavement | bmp-perv-001 through 005 | 4 PASS, 1 WEAK (bmp-perv-003: 0.5 in/hr minimum not re-stated in 2026 §9.6 body text; supported via worked example and Ch.12 reference only) |
| dry well | bmp-drywell-001 through 004 | 3 PASS, 1 FAIL (see below) |
| infiltration basin | bmp-infbasin-001 through 004 | All PASS. 0.5 in/hr, 2-ft SHWT, FOS=2 all confirmed both eras. |
| green roof | bmp-greenroof-001 through 004 | All PASS. Recharge exclusion confirmed; ET framework in Ch.14 confirmed. |
| cistern | bmp-cistern-001 through 004 | 3 PASS, 1 WEAK (bmp-cistern-003: Ch.14 linkage is 2026-era only; claim implies cross-era continuity that doesn't exist) |
⚠️ FAIL: bmp-drywell-003 — Unsourced foundation setback¶
Claim text: "The live dry-well source set keeps the maximum drawdown time at 72 hours and the minimum foundation setback at 10 feet."
Finding: The 72-hour drawdown is confirmed. The "10-foot minimum foundation setback" has no basis in BMP Manual Chapter 9.2 (either the 2023 or 2026 edition). The only "10 feet" in the source documents is the SHWT depth in a worked example. No setback from foundations or buildings is specified in the dry-well chapter. This value also appears in dry-well.md as a design criteria table entry ("Foundation setback | 10 feet minimum") and must be removed or sourced to an actual chapter citation.
Action required: Remove or source the foundation setback value before publication.
Tier 2 — 5 BMPs (15 claims)¶
| Claim ID | Page | Verdict | Notes |
|---|---|---|---|
| bmp-wp10-001 | wet-pond | PASS | Ch.10.5 GI wet pond confirmed |
| bmp-wp10-002 | wet-pond | PASS | Dual Ch.10.5 + Ch.11.6 architecture confirmed |
| bmp-wp10-003 | wet-pond | WEAK | "Embankment condition" not in Ch.10.5 maintenance text; source says "structural components" |
| bmp-cw-001 | constructed-wetland | PASS | Ch.10.4 Standard Constructed Wetlands confirmed |
| bmp-cw-002 | constructed-wetland | PASS | Ch.12/13 chain defensible; Ch.13 reference is less explicit for CW than for infiltration BMPs |
| bmp-cw-003 | constructed-wetland | WEAK | "Invasive-species control" not in source; actual text says "unwanted growth" |
| bmp-sf10-001 | sand-filter | PASS | Ch.10.3 GI sand filter confirmed |
| bmp-sf10-002 | sand-filter | PASS | Ch.12 chain explicitly confirmed in source |
| bmp-sf10-003 | sand-filter | PASS | Pretreatment, media, drawdown all confirmed |
| bmp-br11-001 | blue-roof | PASS | Ch.11.1 Non-GI confirmed |
| bmp-br11-002 | blue-roof | PASS | Non-GI, no recharge, storage-and-release confirmed verbatim |
| bmp-br11-003 | blue-roof | PASS | Drain controls, overflow protection, roof-condition monitoring confirmed |
| bmp-ed11-001 | extended-detention | PASS | Ch.11.2 Non-GI confirmed |
| bmp-ed11-002 | extended-detention | PASS | No-infiltration/no-recharge confirmed verbatim |
| bmp-ed11-003 | extended-detention | WEAK | "Embankment condition" not in Ch.11.2 maintenance text |
Tier 3 — 18 manifests (structural check)¶
| Manifest | Verdict | Notes |
|---|---|---|
| ch9/grass-swale | PASS | Side-slope quote verified verbatim |
| ch9/small-scale-sand-filter | FAIL | bmp-sssf-002 states standard config is "underdrained" — source explicitly excludes underdrained type from Ch.9.9; standard is the infiltrating GI type |
| ch9/vegetated-filter-strip | WEAK | "level-spread" not in 2026 source; correct term is "sheet flow" |
| ch9/rain-garden | PASS | Ch.9.7 derivation, 2.5-acre limit confirmed |
| ch9/tree-trench | PASS | Ch.9.7/9.8 derivation correct |
| ch9/subsurface-gravel-wetland | PASS | Correctly identifies Ch.11.5 Non-GI; taxonomy caution explicit |
| ch10/index | WEAK | Ch.10.1 (Large-Scale Bioretention) exists in corpus but omitted from enrolled family without explanation |
| ch10/extended-detention | PASS | Correctly routes to Ch.11.2 Non-GI |
| ch10/infiltration-trench | PASS | Source-bundle mismatch acknowledged in meta notes |
| ch11/index | WEAK | Ch.11.4 (Sand Filters with Underdrain) and Ch.11.6 unenrolled without explanation |
| ch11/dry-detention | PASS | Ch.11.2 derivation correct |
| ch11/underground-detention | PASS | Ch.11.2 derivation correct |
| ch11/hydrodynamic-separator | PASS | HDS device label confirmed in Ch.11.3 |
| ch11/manufactured-treatment-device | PASS | Ch.11.3 confirmed |
| ch11/membrane-system | WEAK | "Membrane" label absent from Ch.11.3 text; only HDS and filtration device types named |
| ch11/oil-grit-separator | WEAK | "Oil-grit separator" absent from Ch.11.3 text |
| bmp-library/index | PASS | Rain garden/tree trench derivation correct |
| ch9/index | PASS | Derived entries anchored to Ch.9.7/9.8 |
Phase 2D: Design-Workflow (20 claims)¶
All 20 claims PASS. All NJAC section references (7:8-5.3(b), 7:8-5.5(d), 7:8-5.6(b)3, 7:8-5.6(d), 7:8-5.7, 7:8-5.8), BMP chapter references (Ch.6, 8, 12, 13, 14), and process-sequencing assertions confirmed against sources in earlier phases.
Phase 2E: Glossary (6 claims)¶
All 6 claims PASS. WQDS, GI definition, major development, volumetric reduction, underdrain distinction, and Ksat/Ch.12/Ch.13 framework all confirmed against NJAC and BMP Manual sources.
Phase 3: Cross-Module Consistency¶
Spot-checked key values used across multiple modules:
| Value | regulatory-explorer | engineering-constraints | bmp-library | design-workflow | glossary | Consistent? |
|---|---|---|---|---|---|---|
| WQDS = 1.25 in / 2 hr | reg-wq-001 ✓ | — | bmp-bio-002 ✓ | dw-hydro-001 ✓ | glossary-003 ✓ | YES |
| 2-ft SHWT separation | — | eng-shwt-003 ✓ | bmp-drywell-002 ✓ | — | — | YES |
| K_design = K_field / 2.0 | — | eng-soil-002 ✓ | bmp-infbasin-003 ✓ | dw-size-004 ✓ | glossary-006 ✓ | YES |
| 2.5-acre bioretention limit | reg-quantity-001 (Table 5-1) ✓ | eng-da-001 ✓ | bmp-bio-001 ✓ | — | — | YES |
| 7:8-5.3(h) hydraulic-impact trigger | — | eng-mound-001 ✓ | bmp-drywell-004 ✓ | dw-site-003 ✓ | — | YES |
| GI definition (3-part) | reg-gi-001 ✓ | — | — | — | glossary-002 ✓ | YES |
No cross-module inconsistencies found.
Phase 4: 2026 Change Verification¶
Claims asserting something is "new in 2026," "changed," or "unchanged" between eras:
| Claim | Assertion | Verified? |
|---|---|---|
| reg-wq-003 | 7:8-5.6(d) is new in 2026 | CONFIRMED — absent from 2023 NJAC source |
| reg-quantity-002 | 7:8-5.6(d) not in 2023 | CONFIRMED |
| reg-major-002 | Motor vehicle surface and reconstruction are "2026 additions" | PARTIALLY INCORRECT — motor vehicle surface dates to March 2021 (in 2023 rules); only reconstruction trigger is genuinely new in 2026 |
| bmp-bio-001 | 2.5-acre bioretention limit unchanged 2023→2026 | CONFIRMED |
| eng-soil-001 | ASTM D3385 accepted (not D5126) | CONFIRMED in both eras |
| bmp-drywell-002 | 0.5 in/hr and 2-ft SHWT unchanged | CONFIRMED in both era criteria tables |
| bmp-infbasin-002/003 | Permeability and FOS unchanged | CONFIRMED in both era criteria tables |
| bmp-greenroof-002 | 6-inch growing medium limit unchanged | CONFIRMED identical text both eras |
| glossary-004 | Volumetric reduction codified in 7:8-5.6(d) (2026) | CONFIRMED — Ch.14 and 7:8-5.6(d) both new in 2026 |
Action Items — Required Before Publication¶
🔴 FAIL (must fix)¶
1. dry-well.md line 24 — Remove unsourced foundation setback
- Current: Foundation setback | 10 feet minimum
- Finding: No setback requirement appears in BMP Manual Ch.9.2 (either edition). The "10 feet" in the source is a worked-example SHWT depth value.
- Fix: Remove the row from the design criteria table, or locate and cite the actual source if a setback requirement exists elsewhere (e.g., local ordinance or other chapter).
- Also update bmp-drywell-003 claim text to remove the setback assertion.
2. ch9/small-scale-sand-filter — bmp-sssf-002 inverts the source
- Current claim: "The standard small-scale sand-filter configuration...is an underdrained treatment BMP."
- Finding: BMP Manual Ch.9.9 explicitly states "the underdrained type does not meet the definition of green infrastructure...and is not included in this chapter." The Ch.9.9 canonical configuration is the infiltrating GI type; underdrained belongs to Ch.11.4 (unenrolled).
- Fix: Rewrite bmp-sssf-002 to state the standard Ch.9.9 configuration is the infiltrating (GI) type. Note Ch.11.4 for the underdrained variant.
🟡 WEAK (review and resolve)¶
3. reg-major-002 — Motor vehicle surface is not a 2026 addition
- The motor vehicle surface trigger has been in the NJAC since March 2, 2021 (present in the 2023 rules). Only the reconstruction trigger (item 4) and the expanded combination provision (item 5) are genuinely new in 2026.
- Fix: Revise the claim text to separate the two triggers: note motor vehicle surface as a 2021-era threshold that remains in 2026, and reconstruction as the genuine 2026 addition.
4. bmp-perv-003 — 0.5 in/hr minimum not re-stated in 2026 §9.6 infiltrating-type body text
- The threshold is supported by the Ch.12 cross-reference and worked examples, but the 2026 §9.6 does not restate it explicitly in the infiltrating-type design section as the 2023 version did.
- Fix: Add a Ch.12 cross-reference to the claim source_refs to make the evidentiary chain explicit.
5. bmp-cistern-003 — Chapter 14 linkage framed as cross-era
- Ch.14 is new in 2026; the 2023 source has no equivalent. The claim's framing implies a continuing framework across both eras.
- Fix: Reframe as "in the 2026 source set, cisterns are placed in the volumetric-reduction framework through Ch.9.1 and new Ch.14."
6. bmp-wp10-003 / bmp-ed11-003 — "Embankment condition" not in source maintenance text
- The BMP Manual uses "structural components" generically; it does not name "embankment" as a discrete maintenance category for wet ponds or extended detention basins.
- Fix: Replace "embankment condition" with "structural component condition" to match source language.
7. bmp-cw-003 — "Invasive-species control" not in source
- Source says "unwanted growth should be removed," not invasive species.
- Fix: Replace "invasive-species control" with "unwanted growth removal."
8. ch11/membrane-system and ch11/oil-grit-separator — Labels absent from Ch.11.3
- The terms "membrane system" and "oil-grit separator" do not appear in the Ch.11.3 source text.
- Fix: Add a note to each manifest acknowledging these are industry-conventional labels not present in the cited chapter text. Consider cross-referencing the NJDEP MTD certification list.
9. ch10/index — Ch.10.1 omission unexplained; ch11/index — Ch.11.4 and Ch.11.6 omission unexplained
- These chapters exist in the corpus and are cross-referenced in enrolled chapters' body text.
- Fix: Add a note to each index manifest explaining why these chapters are not enrolled (e.g., enrollment deferred, outside scope of this module, etc.).
Notes on Reports Module¶
The 8 legacy reports (inputs/Reports/) have no *.claims.json manifests. These were identified as out of scope for the current validation pass (they are synthesis/working documents, not primary published content). The reports should not be treated as authoritative source — the 5 primary modules are the validated content layer.
If any Report content is surfaced directly to end users, it should be reviewed against the validated modules before display.
Validation Status by Module¶
| Module | Ready to publish? | Conditions |
|---|---|---|
| regulatory-explorer | ✅ YES | Address WEAK on reg-major-002 (not blocking) |
| engineering-constraints | ✅ YES | Clean |
| bmp-library — bioretention | ✅ YES | Clean |
| bmp-library — pervious pavement | ✅ YES | Address bmp-perv-003 citation (not blocking) |
| bmp-library — dry well | ❌ NO | Remove unsourced 10-ft foundation setback first |
| bmp-library — infiltration basin | ✅ YES | Clean |
| bmp-library — green roof | ✅ YES | Clean |
| bmp-library — cistern | ✅ YES | Address bmp-cistern-003 era framing (not blocking) |
| bmp-library — wet pond | ✅ YES | Address embankment language (not blocking) |
| bmp-library — constructed wetland | ✅ YES | Address unwanted growth language (not blocking) |
| bmp-library — sand filter (Ch.10.3) | ✅ YES | Clean |
| bmp-library — blue roof | ✅ YES | Clean |
| bmp-library — extended detention | ✅ YES | Address embankment language (not blocking) |
| bmp-library — small-scale sand filter | ❌ NO | Fix inverted GI/underdrain classification claim first |
| bmp-library — derived entries (rain garden, tree trench, etc.) | ✅ YES | Clean |
| bmp-library — membrane system, oil-grit separator | ⚠️ CONDITIONAL | Labels unverified against source; add note |
| design-workflow | ✅ YES | Clean |
| glossary | ✅ YES | Clean |
Summary¶
-Summary' 137 claims verified across 48 manifests. 127 PASS, 9 WEAK, 1 FAIL (corrected to 2 after cross-checking).
Must fix before publishing (🔴) dry-well.md line 24 — Remove "Foundation setback | 10 feet minimum." This value has no basis in BMP Manual Ch.9.2. The "10 feet" in the source is an example SHWT depth in a worked problem.
small-scale-sand-filter bmp-sssf-002 — Inverted classification. Ch.9.9 is the infiltrating (GI) type; the source explicitly says the underdrained type is excluded from this chapter. Rewrite to reflect this.
Review before publishing (🟡) reg-major-002 — Motor vehicle surface trigger dates to 2021, not 2026. Only the reconstruction trigger is genuinely new. bmp-perv-003 — 0.5 in/hr minimum not re-stated in 2026 §9.6 body text; cite Ch.12 explicitly. bmp-cistern-003 — Ch.14 linkage is 2026-era only; claim implies cross-era continuity. bmp-wp10-003 / bmp-ed11-003 — Replace "embankment condition" with "structural component condition" (source language). bmp-cw-003 — Replace "invasive-species control" with "unwanted growth removal" (source language). Membrane-system / oil-grit-separator — Labels absent from Ch.11.3 text; add a note acknowledging they're industry-conventional, not chapter-defined. ch10/index / ch11/index — Add notes explaining why Ch.10.1, Ch.11.4, and Ch.11.6 are unenrolled. No corrections needed — these are clean