What we are asserting, and how firmly.

Every claim below carries a confidence score and, where the claim has legal weight, a Legal Confidence Score (LCS). Low-confidence claims are not hidden. They sit alongside the high-confidence ones so a reader can tell the difference.

Gold Star Green claims with confidence and sources
StatementCategoryConfidenceLCSSource
The City of Cambridge Department of Public Works has published two primary soil-testing tables and one groundwater summary table for Gold Star Mothers Park. The numeric values in the table below are re-extracted verbatim from those PDFs as of 2026-04-13.contamination90%L3primary (kb/gold-star-green/raw/pdfs/table-1-initial-release.pdf)
As of 2026-04-13, no licensed structural, geotechnical, or environmental engineer has produced a report evaluating the park-roof structural concept at Gold Star Mothers Park. Every engineering sub-claim below is a concept-level assertion, bounded by confidence at or below 0.5 per spec §14.4.engineering40%credible-secondary (kb/gold-star-green/wiki/concepts/park-roof-typology.md)
As of 2026-04-13, no remediation method has been authorized for Gold Star Mothers Park under 40 CFR 761.61(c) or any other pathway. The claims enumerated below describe the regulatory posture verbatim from regulation text and from internal research memos, each with an LCS level.regulatory70%L3primary

Claim details

Contamination Data — Gold Star Mothers Park

Category
contamination
Confidence
90%
LCS
L3

Verified against

  • kb/gold-star-green/raw/pdfs/table-1-initial-release.pdf
  • kb/gold-star-green/raw/pdfs/table-2-soil-testing.pdf

Contamination Data — Gold Star Mothers Park

Every numeric value in this article was re-extracted from the source PDFs listed in verified-against: on 2026-04-13 by agent A4. Values were not copied from TSX pages, SQL seeds, or earlier working notes. Where those prior artifacts disagreed with the PDF, the PDF value is preserved and the discrepancy is flagged in the “Discrepancies found during re-extraction” section below.

Primary sources, verbatim headings:

  • kb/gold-star-green/raw/pdfs/table-1-initial-release.pdf — “Table 1, Release Notification Form Table, Gold Star Mothers Park, Cambridge MA”
  • kb/gold-star-green/raw/pdfs/table-2-soil-testing.pdf — “Table 2, Gold Star Mothers Park, Cambridge MA, Additional Soil Testing Data”

Both tables record Oil or Hazardous Materials (O or HM) with CAS numbers, concentrations in mg/kg, and the Reportable Concentration (RCS-1) threshold for comparison. RCS-1 is the Massachusetts Contingency Plan threshold that requires reporting to MassDEP.

Table 1 reports one value per contaminant (column: “Amount or Concentration”). Table 2 reports both a Maximum Detected Concentration and an Average Detected Concentration.

Table 1 — Release Notification Form (verbatim re-extraction)

Source: kb/gold-star-green/raw/pdfs/table-1-initial-release.pdf (page 1).

ContaminantCASAmount or ConcentrationUnitRCs Exceeded
Antimony7440-36-0935MG/KGRCS-1
Arsenic7440-38-2156MG/KGRCS-1
Barium7440-39-33,120MG/KGRCS-1
Chromium (Total)7440-47-3127MG/KGRCS-1
Lead7439-92-14,420MG/KGRCS-1
Zinc7440-66-62,730MG/KGRCS-1
PCBs, Total~37.9MG/KGRCS-1
Benzo(a)anthracene56-55-337MG/KGRCS-1
Benzo(a)pyrene50-32-837MG/KGRCS-1
Benzo(b)fluoranthene205-99-240MG/KGRCS-1
Dibenz(a,h)anthracene53-70-34.2MG/KGRCS-1
Indeno(1,2,3-CD)pyrene193-39-523MG/KGRCS-1
2-Methylnaphthalene91-57-60.87MG/KGRCS-1
Phenanthrene85-01-841MG/KGRCS-1

All Table 1 values come from a single column labelled “Amount or Concentration.” The table does not distinguish maximum versus average. Interpretation: these are the reported values used for the release notification filing under the Massachusetts Contingency Plan.

Table 2 — Additional Soil Testing Data (verbatim re-extraction)

Source: kb/gold-star-green/raw/pdfs/table-2-soil-testing.pdf (page 1).

ContaminantCASMaximum DetectedAverage DetectedRCS-1Unit
Arsenic7440-38-2100.012.020MG/KG
Barium7440-39-36100.0343.01000MG/KG
Cadmium7440-43-9210.013.480MG/KG
Chromium7440-47-3380.029.6100MG/KG
Lead7439-92-18200.0512.6200MG/KG
Zinc7440-66-61600.0301.61000MG/KG
PCBs, Total1336-36-368.04.81MG/KG
2-Methylnaphthalene91-57-648.02.90.7MG/KG
Acenaphthene83-32-997.04.14MG/KG
Acenaphthylene208-96-812.01.52MG/KG
Benzo(a)anthracene56-55-3510.015.220MG/KG
Benzo(a)pyrene50-32-8350.012.12MG/KG
Benzo(b)fluoranthene205-99-2390.012.920MG/KG
Chrysene218-01-9470.014.6200MG/KG
Dibenz(a,h)anthracene53-70-3100.03.82MG/KG
Fluoranthene206-44-01100.028.61000MG/KG
Indeno(1,2,3-CD)pyrene193-39-5240.07.920MG/KG
Naphthalene91-20-397.04.04MG/KG
Phenanthrene85-01-81100.026.610MG/KG

Column headings verbatim from the PDF: “O or HM Released”, “CAS Number, if known”, “O or HM”, “Maximum Detected Concentration”, “Average Detected Concentration”, “RCS-1 Concentration”, “Units”. Values reported in MG/KG (“Oil or Hazardous Material” = “HM” for every row). The footer of the PDF defines RCS-1 as “threshold that determines when a contaminant must be reported to MassDEP.”

Derived multiples (computed from the PDF values only)

These multiples are computed from the PDF values above divided by the RCS-1 threshold published in the same table. They are not transcribed from any other source.

  • Lead maximum: 8200.0 / 200 = 41x RCS-1 (Table 2).
  • PCBs Total maximum: 68.0 / 1 = 68x RCS-1 (Table 2).
  • Benzo(a)pyrene maximum: 350.0 / 2 = 175x RCS-1 (Table 2).
  • Dibenz(a,h)anthracene maximum: 100.0 / 2 = 50x RCS-1 (Table 2).
  • Phenanthrene maximum: 1100.0 / 10 = 110x RCS-1 (Table 2).
  • Benzo(a)anthracene maximum: 510.0 / 20 = 25.5x RCS-1 (Table 2).
  • Fluoranthene maximum: 1100.0 / 1000 = 1.1x RCS-1 (Table 2).

Discrepancies found during re-extraction (CRITICAL)

Prior working documents under raw/extracts/bioremediation-initiative-seed.md (SQL seed) and the TSX page extracts reported some values that do NOT match the PDF. Per the do-not-strengthen-claims invariant (§22) and the verify-transcribed-numbers rule, the PDF is treated as the source of truth and the discrepancies are preserved here rather than silently corrected.

  1. PCBs Total maximum. Seed says 48 ppm peak; Table 2 PDF says 68.0 mg/kg. Discrepancy of +20 mg/kg. PDF preserved.
  2. Benzo(a)pyrene maximum. Seed says 390 ppm peak; Table 2 PDF says 350.0 mg/kg. Discrepancy of -40 mg/kg. PDF preserved. The 390 figure may be a transposition of the benzo(b)fluoranthene max (390.0).
  3. Dibenz(a,h)anthracene maximum. Seed says 1,100 ppm peak; Table 2 PDF says 100.0 mg/kg. Order-of-magnitude discrepancy. PDF preserved. The 1,100 figure matches phenanthrene and fluoranthene in Table 2 and may have been a row misread.
  4. Acenaphthylene maximum. Seed says 510 ppm peak; Table 2 PDF says 12.0 mg/kg. Order-of-magnitude discrepancy. PDF preserved. The 510 figure matches benzo(a)anthracene max (510.0) in Table 2 and may have been a row misread.
  5. Lead average. The seed’s aggregate claim of “lead at 8,200 ppm — 41x EPA residential screening level” omits the average value; the Table 2 PDF records 512.6 mg/kg average. The PDF’s own published threshold in the same table is RCS-1 = 200 mg/kg (MassDEP), not the EPA residential screening level of 200 mg/kg that the seed references. Both thresholds coincide numerically at 200 mg/kg but derive from different regulations. The language used here cites only the published RCS-1 threshold from the same PDF.

No downstream article in wiki/ should use the seed values for items 1-4. Articles that need a contamination number should reference this article and use the re-extracted value.

What the PDFs do not state

  • The PDFs do not attribute a source contaminant to a particular historical event. Attribution to the 1963 Squire’s meatpacking fire appears in secondary sources (community archive, bioremediation research memos) and is treated as community or credible-secondary in other articles, not as a Table 1 or Table 2 claim.
  • The PDFs do not state a depth for each sample. Depth information, if any, would be in supplementary sampling narratives not re-extracted into this article.
  • The PDFs do not state which samples came from which grid cells. Spatial distribution, if needed, should be re-extracted from the community meeting slide deck at raw/pdfs/community-meeting-presentation-oct16.pdf rather than inferred from the tables.
  • The PDFs do not state the analytical method or laboratory per contaminant.

Confidence and caveats

  • Overall confidence: 0.9 reflects that the values here are direct re-extractions from the primary source PDFs and that the PDF text is legible and unambiguous. The confidence is not 1.0 because: (a) the PDFs themselves are derived from sampling and lab work performed by or for the City, which introduces its own measurement uncertainty that is not reported in the PDFs; (b) no independent lab verification has been performed; (c) Dibenz(a,h)anthracene on Table 2 reads as 100.0 — the value is legible but large numerical discrepancies against seeded working docs are preserved in the section above, so downstream consumers should cross-check before using.
  • lcs: 3 — regulatory/reporting posture is informational only; this article is not a legal filing and is not counsel-reviewed.
  • gold star mothers park — the case file.
  • observable land — the concept that governs how this article is maintained.
  • regulatory pathway — the regulatory claims built on this data.
  • engineering feasibility — the engineering claims that depend on these values.

Engineering Feasibility — Park-Roof at Gold Star Mothers Park

Category
engineering
Confidence
40%

Engineering Feasibility — Park-Roof at Gold Star Mothers Park

Standing caveat, repeated per sub-claim: no licensed structural, geotechnical, or environmental engineer has produced a report on record for the Gold Star Mothers Park park-roof concept as of 2026-04-13. Every sub-claim in this article is concept-level and bounded at confidence ≤ 0.5. Spec §14.4 requires this bound. Nothing here should be read as a statement that a specific structure is buildable.

Sub-claim 1 — Park-roof is a real architectural typology with large-scale precedents

  • Statement: “Park-roof, meaning a public park surface supported by a structural system above a use that cannot be opened to the public (active rail, sunken highway, contaminated fill, etc.), has been built at regional scale in multiple US cities.”
  • Confidence: 0.5 — ceiling for the engineering category.
  • Source tier: credible-secondary (precedent articles and public project documentation).
  • Notes: Examples include Millennium Park (Chicago, 2004), Klyde Warren Park (Dallas, 2012), and The High Line (New York, 2009). See the respective precedent articles. This sub-claim asserts precedent existence only — it does not assert that any of these precedents is a direct analog for Gold Star Mothers Park.

Sub-claim 2 — A park-roof is compatible with different treatment scenarios below the slab

  • Statement: “Conceptually, a park-roof system can be designed to accommodate either a fully excavated sub-grade or an intact-fill sub-grade with an in-situ treatment zone. Which one is actually built depends on the City’s remediation decision and on detailed geotechnical, structural, and regulatory review.”
  • Confidence: 0.4.
  • Source tier: credible-secondary (internal research memos, precedent articles).
  • Notes: This sub-claim is weaker than sub-claim 1 because it asserts design-space scope, not design correctness. No specific section, load path, or foundation system has been engineered for Gold Star.

Sub-claim 3 — Helical pile foundations are a known option for load-bearing structures over soft or filled ground

  • Statement: “Helical piles are a known foundation technology used under structures where soft fill or contamination complicates shallow footings. Their use at a specific site depends on site-specific geotechnical investigation.”
  • Confidence: 0.4.
  • Source tier: credible-secondary (engineering practice documentation in general; not a Gold Star-specific report).
  • Notes: No geotechnical investigation has been performed for Gold Star in the context of a park-roof. The Boston Blue Clay layer mentioned in the groundwater summary does not by itself determine a foundation system.

Sub-claim 4 — Loads for a public park deck are calculable but not calculated here

  • Statement: “Structural loads for a public park deck — including dead loads, live loads, tree and soil loads, wind, and seismic — are standard engineering calculations. They have not been performed for Gold Star Mothers Park as of 2026-04-13.”
  • Confidence: 0.5.
  • Source tier: credible-secondary.
  • Notes: The purpose of this sub-claim is to prevent any user-facing page from implying that loads have been analyzed. They have not.

Sub-claim 5 — A treatment chamber below a slab would add unusual structural and ventilation requirements

  • Statement: “If an in-situ treatment chamber were placed below a park-roof slab, the resulting system would have unusual requirements compared to a conventional park-over-infrastructure project: managed atmosphere, access for monitoring and maintenance, emergency venting and containment, and long-term integrity verification. None of these have been specified for Gold Star.”
  • Confidence: 0.3 — deliberately low because the concept is novel and there is no analog project on record.
  • Source tier: credible-secondary (research memos).
  • Notes: This sub-claim is the one most likely to be contested on review. It is stated as an enumeration of requirements rather than as a design solution.

Sub-claim 6 — The cross-section SVG on the site is a conceptual illustration, not an engineered drawing

  • Statement: “Any cross-section diagram used in the site or in partner materials is a conceptual illustration. It is not an engineered drawing, is not to scale, and should not be treated as a structural proposal.”
  • Confidence: 0.5.
  • Source tier: credible-secondary (Spec §22 visual invariant; cross-section SVG owned by agent A8).
  • Notes: Spec §22 requires that the cross-section SVG carry DOM text stating “Conceptual section. Not to scale. Not an engineered drawing.” Any party consuming this article in a user-facing surface must preserve that text.

Sub-claim 7 — Treatability and scale-up for bioremediation-based approaches at this site are unestablished

  • Statement: “Bench-scale or pilot-scale treatability data using Gold Star Mothers Park soils is not in the public record as of 2026-04-13. Any field-scale deployment of a remediation approach at this site would need such data, and the 40 CFR 761.61(c) application skeleton in the raw archive identifies a treatability study as the single most important piece of technical evidence.”
  • Confidence: 0.5.
  • Source tier: credible-secondary (application skeleton memo, section discussing Gate 4 and the treatability study requirement).
  • Notes: This sub-claim is a bound on what the project can currently claim. It deliberately does not name a specific bioremediation method.

Overall engineering posture summary

  • No licensed engineer on file.
  • No geotechnical report on file.
  • No structural drawings on file.
  • No treatability report on file.
  • No cost estimate is treated as engineering-grade; the $10M excavation estimate publicly attributed to the City belongs to the City’s process, not to the project.

Any forward statement of “we could build this” in user-facing surfaces must be qualified as conceptual. Qualified language is required until a licensed engineer is engaged.

  • park roof typology — the concept article.
  • contamination data — the data that any engineering review would consume.
  • gold star mothers park — the case file.
  • cambridge coalition graph — partners who could be engaged to produce a qualified engineering review.

Regulatory Pathway — Gold Star Mothers Park

Category
regulatory
Confidence
70%
LCS
L3

Regulatory Pathway — Gold Star Mothers Park

This article enumerates regulatory sub-claims. Each sub-claim is stated, labelled with an LCS level per memory/legal_confidence_system.md, and sourced. Default for any sub-claim that has not been counsel-reviewed is lcs: 3 (or lower). No sub-claim below has been counsel-reviewed as of 2026-04-13; every LCS value is therefore conservative.

The overall article carries confidence: 0.7 because regulatory interpretation is a matter of discretion and because some sub-claims depend on a working-draft research memo rather than on a filed document.

Sub-claim 1 — 40 CFR 761.61 exists and provides three pathways

  • Statement: “40 CFR 761.61 provides three pathways for PCB remediation waste: self-implementing under (a), performance-based under (b), and risk-based approval under (c).”
  • Source: primary (eCFR, 40 CFR 761.61).
  • LCS: 6 — this is direct reading of the regulation. Confident, but labelled as informational because regulation text is not a legal opinion.
  • Confidence: 0.95.
  • Notes: This sub-claim restates regulation text. It is not a claim that any specific application has been approved or that any specific pathway applies to Gold Star Mothers Park.

Sub-claim 2 — 761.61(c) is a discretionary risk-based approval pathway

  • Statement: “Under 40 CFR 761.61(c), any person wishing to clean up or dispose of PCB remediation waste by a method other than those prescribed in (a) or (b) must apply in writing to the Regional Administrator and receive written approval. The approval standard is ‘unreasonable risk of injury to health or the environment.’”
  • Source: primary (eCFR, 40 CFR 761.61(c)).
  • LCS: 5 — direct quote from the regulation. Lowered from 6 because the practical content of “unreasonable risk” is discretionary and case-specific.
  • Confidence: 0.9.
  • Notes: Approval is discretionary, not guaranteed. The regulation does not set a shot clock on a 761.61(c) decision.

Sub-claim 3 — Self-implementing (a) and performance-based (b) pathways constrain high-occupancy sites

  • Statement: “For high-occupancy areas such as residential sites, parks, playgrounds, schools, and day care centers, the 761.61(a) and 761.61(b) pathways include specific cleanup-level constraints, and 761.61(b)(1) is expressly not available at some categories of high-occupancy sites.”
  • Source: primary (eCFR, 40 CFR 761.61(a) and (b)) plus credible-secondary paraphrase from the application skeleton memo.
  • LCS: 4 — the memo’s reading is plausible but has not been counsel-reviewed, and the precise interplay between (a) and (b) at a park site depends on facts an application would need to document.
  • Confidence: 0.75.
  • Notes: This sub-claim is the premise for concluding that a bioremediation-based approach at Gold Star Mothers Park would proceed under (c). The premise is weaker than the conclusion; see the do-not-strengthen-claims invariant.

Sub-claim 4 — PCB authority under TSCA Section 6(e) is non-delegable

  • Statement: “PCBs are regulated under TSCA Section 6(e); this authority is retained by federal EPA and is not delegated to states, so Massachusetts cannot issue a state-level PCB cleanup approval.”
  • Source: credible-secondary (application skeleton memo and deep-dive memo).
  • LCS: 3 — correctly described as a stable feature of TSCA in multiple research memos, but this article does not cite the statute directly for non-delegability.
  • Confidence: 0.75.
  • Notes: MassDEP retains authority over the non-PCB fractions of the contamination (lead, PAHs, metals) and over the Massachusetts Contingency Plan (MCP). The MCP regime is a parallel and independent obligation.

Sub-claim 5 — EPA Region 1 is the Regional Administrator for any 761.61(c) application at this site

  • Statement: “An application under 40 CFR 761.61(c) for a site in Cambridge, Massachusetts would be submitted to the Regional Administrator of EPA Region 1 (New England).”
  • Source: primary (EPA Region 1 public information) plus regulation text.
  • LCS: 5.
  • Confidence: 0.95.
  • Notes: This is a jurisdictional fact, not a claim of engagement. See epa region 1 for the entity stub.

Sub-claim 6 — No 761.61(c) application has been filed for Gold Star Mothers Park

  • Statement: “As of 2026-04-13, there is no filed 40 CFR 761.61(c) application for Gold Star Mothers Park.”
  • Source: credible-secondary (sprints log, santaprosper diff pack, internal knowledge).
  • LCS: 3 — absence of a public notification or record cannot be definitively proven from inside the KB; the statement is based on the absence of any such document in the raw archive and on the April 12 diff pack’s explicit note that “no remediation authorized; no 761.61(c) filed.”
  • Confidence: 0.9.
  • Notes: This sub-claim is important as a guard against implied-approval framing anywhere else in the build.

Sub-claim 7 — No known precedent exists for a 761.61(c) approval using bioremediation as the primary PCB remedy

  • Statement: “No known precedent exists for a 40 CFR 761.61(c) approval that uses bioremediation as the primary remedy for PCBs in soil. Prior 761.61(c) approvals (Parker Street / McCoy Field in New Bedford, MA, 2005) have been for conventional excavation with institutional controls.”
  • Source: credible-secondary (application skeleton memo at lines 548-555 of raw/moonshots.boston/research/homeworld-bio/761-61c-application-skeleton.md).
  • LCS: 3 — an absence-of-precedent claim is weak by nature; the memo is a working draft.
  • Confidence: 0.7.
  • Notes: Bioremediation of PCBs has been evaluated under EPA SITE and DoD ESTCP research programs; it has not been approved as a standalone remedy under TSCA in the known record. A Gold Star 761.61(c) application using bioremediation would be first-of-kind on that axis.

Sub-claim 8 — Massachusetts Contingency Plan RTNs 3-52413 and 3-52198 apply to this site

  • Statement: “The Massachusetts Contingency Plan (MCP) assigns Release Tracking Numbers (RTNs) to reported releases at Gold Star Mothers Park: RTN 3-52413 (overall park) and RTN 3-52198 (basketball court).”
  • Source: credible-secondary (bioremediation initiative seed; application skeleton memo).
  • LCS: 3 — unverified against the MassDEP RTN lookup in this compilation pass.
  • Confidence: 0.8.
  • Notes: MCP obligations proceed under MassDEP regardless of the TSCA posture; they govern Chapter 21E liability, Licensed Site Professional involvement, and deadlines for response actions.

Sub-claim 9 — The site is a public park and a sensitive-receptor location

  • Statement: “Gold Star Mothers Park is a public park dedicated by the City of Cambridge and is therefore a high-occupancy, sensitive-receptor location under the 761.61 framing.”
  • Source: primary (City of Cambridge park user guide raw/pdfs/park-user-guide.pdf and dedication language) and the regulation’s treatment of parks.
  • LCS: 4.
  • Confidence: 0.9.
  • Notes: The sensitive-receptor framing is why (a) and (b) are constrained and why (c) is the relevant pathway for any alternative method. See gold star mothers park.

Sub-claim 10 — The City has appropriated funds and estimated total cost, which is not itself a regulatory approval

  • Statement: “The City of Cambridge has appropriated funds for remediation work at Gold Star Mothers Park and has publicly described a higher total estimated cost (see contamination data for the data source trail). Appropriation is not a regulatory approval under TSCA or MCP.”
  • Source: credible-secondary (bioremediation seed memo, April 12 diff pack).
  • LCS: 3 — figures are not re-extracted from a primary city council PDF in this compilation pass.
  • Confidence: 0.7.
  • Notes: The specific dollar figures are omitted from this article and deferred to a future pass that re-extracts them from a council filing. Do not cite the seed’s dollar figures as primary-sourced.

Overall regulatory posture summary

Taken together the sub-claims above support a narrow, carefully-worded regulatory posture:

  • The 761.61(c) pathway exists and is the relevant pathway for any alternative-method approach at this site.
  • No application has been filed. No approval has been granted.
  • The City of Cambridge is the property owner and the only party with authority to direct a filing for this site.
  • MassDEP retains jurisdiction over the non-PCB fractions and over MCP obligations in parallel.

Nothing in this article should be read as a claim of EPA, MassDEP, City of Cambridge, or any other regulator’s endorsement, approval, or commitment.

  • 761 61c pathway — the concept article.
  • epa region 1 — the regulatory entity stub.
  • city of cambridge — the property-owner entity stub.
  • contamination data — the re-extracted numeric record.