ABOUT · METHODS RECORD
How this site is made
The current rules, grading scale, checks and records behind every page, generated at each build from the same files the pages are built from.
How a page is made (current process)
Each of the 26 pages is a structured text file with a claims ledger. Every factual claim with a number, a comparison or a strong statement is recorded with its source, scope, year, grade and a one-line reason for the grade. At this build the ledgers hold 877 claims (A 26, B 74, C 375, D 54, E 46, P 302).
Sources come from one shared registry of 429 entries; a page may cite only registry entries. Claims were checked against their sources, pages were reviewed group by group by a second AI model acting as a skeptical scientist, and corrections were applied and recorded. Who approves publication: to be confirmed by the owner.
Rules every page follows
- No claim without a source (DOI or stable URL), year and, for key claims, an evidence level A–E with a one-line rationale. Peer-reviewed first; agency reports second; reputable press third; preprints marked as preprints.
- Numbers carry unit, scope (where, when, which size class, which method). Disagreements between studies are shown, not hidden. Model outputs (the home counters) are "estimates, not measurements".
- No fabrication: no invented studies, numbers, people, quotes, events or images. If something cannot be sourced, say "not established" and grade E.
- Visuals: real photos need license and credit; AI images are always labelled "illustration"; data graphics ship with their data table and source; captions say what is shown and at what scale.
Evidence levels and provenance
- A · Consensus
- Multiple independent studies or a systematic review plus a statement by a scientific body.
- B · Well supported
- Several independent studies with concordant findings.
- C · Limited
- Few studies, narrow conditions, or methodological caveats.
- D · Disputed
- Conflicting results; both sides are shown.
- E · Unknown
- No reliable evidence found; the search scope is stated.
- P · Provenance
- Provenance: checked against the primary record. A separate category, not an evidence grade: used for bibliographic facts, legal texts, official metadata, definitions quoted from a standard, the site's own arithmetic and statements about the site's process.
Grades belong to single findings, not to studies or topics. D and E are not lower rungs of the A–C scale, and P is not a grade.
Automated checks at every build
- Every cited source exists in the registry; every page meets its claims minimum; grades and statuses use the allowed values.
- A list of 79 banned phrases (marketing language, generic attributions such as "studies show") is enforced.
- Images appear only with a recorded credit and license; AI images always carry the "Illustration" label; data graphics ship with their data table.
- Internal links are checked, and the pages are tested for keyboard use and with an automated accessibility checker (axe).
This build (4 November 2026): 26 pages checked, 0 errors, 78 warnings (warnings are editorial notes, such as a source without a publication year).
Downloads
- Sources registry · 429 entries, JSON
- Home counter model · every preset, conversion factor and removed value, JSON
Site records
Registry entries that cite this site's own records. Pages link here as their source.
S-276 · Home counter model of the Nanoplastics planetary observatory: v6 script (NP object) and v7 specification (counter-model.json)
- Date
- 4 November 2026
- Record
- download
S-299 · Dated literature and registry search record for the Human Body page (negative findings)
- Date
- 4 November 2026
- Record
- the record follows
All searches run on 3 November 2026 by the fixer. Europe PMC REST API (/webservices/rest/search, format=json), ClinicalTrials.gov API v2 (/api/v2/studies). Screening: title, and abstract where the title was ambiguous; inclusion = primary human data that would contradict the negative statement.
- S1, dermal uptake (C-human-008). Europe PMC:
(TITLE_ABS:microplastic* OR TITLE_ABS:nanoplastic*) AND (TITLE:dermal OR TITLE:skin OR TITLE:percutaneous OR TITLE:transdermal) AND (PUB_YEAR:[2019 TO 2026]). 34 hits, all screened. None measures particle uptake through intact skin in living people. Studies use skin cells or organoids, 3D skin equivalents (for additive chemicals, not particles), fish or earthworm skin, or sweat-aggregation experiments and models (e.g. doi:10.1016/j.jhazmat.2025.139100; doi:10.1021/acs.est.6c00119). One 2024 hit is about blood contamination from coronary interventions, not skin. - S2, clinical and cohort studies (C-human-030, -041). Europe PMC:
(TITLE:microplastic* OR TITLE:nanoplastic* OR TITLE:"micro- and nanoplastics" OR TITLE:"micro-nanoplastics") AND (TITLE:plaque OR TITLE:atheroma* OR TITLE:carotid OR TITLE:coronary OR TITLE:cardiovascular OR TITLE:stroke OR TITLE:myocardial OR TITLE:arter* OR TITLE:aort* OR TITLE:cohort OR TITLE:prospective OR TITLE:mortality OR TITLE:outcome*) AND (PUB_YEAR:2026 OR PUB_YEAR:2025) NOT PUB_TYPE:review. 92 hits, all screened. Prospective human studies with outcomes: Zhang et al. 2025 (doi:10.1016/j.jhazmat.2025.137624; 142 MI patients, 110 followed 31.5 months, i.e. existing disease); a 2026 bronchoscopy study (cases vs controls at diagnosis); a 2026 gestational-diabetes cohort (doi:10.1007/s12020-026-04708-2; exposure by questionnaire, not body measurement); a 2026 UK Biobank analysis (doi:10.1177/10732748261493417; basin-level river data as an ecological proxy). Cross-sectional: Paolisso 2026, Yang 2026 (letter), femoral and aortic plaque studies 2025–2026. None measures plastic in body samples of disease-free people and follows incident events. - S3, any 2026 clinical micro/nanoplastic study after Marfella (R-planet-25). Europe PMC:
(microplastic OR nanoplastic) AND (plaque OR atheroma OR coronary OR cardiovascular OR stroke OR carotid) AND PUB_YEAR:2026 AND (SRC:MED). 435 hits; first 60 screened (mostly reviews). Result: Paolisso et al. 2026 (added), plus the items listed under S2. A general web search (extended) for the Li 2026 brain study found doi:10.1038/s44360-026-00091-4. - S4, registered studies (C-human-030). ClinicalTrials.gov:
query.term=microplastic OR nanoplastic OR microplastics OR nanoplastics, 51 registrations. Completed observational ones: NCT05900947 (Marfella's carotid study). Prospective outcome studies are registered but not completed, e.g. NCT07748767 (coronary outcomes, not yet recruiting), NCT06107933 (early-life development, recruiting), NCT06316609 (prenatal exposure and atopic dermatitis, active), NCT07140601 (neurological disorders, recruiting). - S5, clinical blood test or reference range (C-human-031). Europe PMC:
(TITLE_ABS:microplastic* OR TITLE_ABS:nanoplastic*) AND (TITLE_ABS:blood OR TITLE_ABS:serum OR TITLE_ABS:plasma) AND ("reference range" OR "reference interval" OR "reference values" OR "clinical test" OR "diagnostic test" OR "clinical assay" OR "biomonitoring guidance value") AND (PUB_YEAR:[2019 TO 2026]). 8 hits, all screened: a flow-cytometry screening method (2023), ICP-TOFMS sizing, reviews; none validates a clinical test or reports a reference interval.
Limits: one database and one registry; English titles; no full-text search. The E grades hold for this scope only.
S-344 · Home page script and markup of the v6 prototype (data model NP, COPY strings, counter and settings code)
- Date
- 4 November 2026
- Record
- kept by the site; summarized on this page
S-345 · Editorial decisions record and complete-solution criterion of the Nanoplastics planetary observatory
- Date
- 4 November 2026
- Record
- kept by the site; summarized on this page
S-390 · Sources registry of the Nanoplastics planetary observatory (the site's internal record)
- Date
- 4 November 2026
- Record
- download
S-392 · Methods and production record of the Nanoplastics planetary observatory (project brief, content schema and tools, agent prompts, registry build notes, verification and review audits)
- Date
- 4 November 2026
- Record
- kept by the site; summarized on this page
Superseded descriptions
The descriptions below come from the project's first drafts and no longer describe the site. They are listed so that older copies can be checked.
- Earlier documents describe the planet-to-cell journey as lasting 17 seconds. The journey as built runs for 22 seconds.
- Until 3 November 2026 the home counters used unsourced low, medium and high scenarios (v6). Since 4 November 2026 (v7) each counter uses a value from one named study; the changes are listed on Flows & Model.