What a Commons Report Is For, and the Rules That Keep It Honest

A record kept by the people caring for each moku, not by the State. Where a report tracks State funding, as the Green Fee report does, the funded projects are listed from the State's own budget worksheet; places, moku, and effects appear as the people doing the work report them. Nothing here is a State report, and no government adoption, endorsement, partnership, or deployment is claimed. The reports support professional review and public accountability; they make no funding, eligibility, policy, or enforcement decisions.

1. What a commons report is for

A commons report is the public side of a record kept in common: who is caring for a place, what they did, what came of it, and where each fact came from. Some reports track public money. The Green Fee report lists every project the Legislature funded with the State's fee on visitors, each in the moku it serves, with every public notice about it recorded with its source and every figure a project reports traceable to who observed it, where, and when. Others will track a land trust's own work, a community group's season, or a watershed's monitoring, with no State funding involved. Two uses follow for every report, and this document exists to keep both honest:

Public accountability. A resident, a steward, a reporter, or a legislator can cite a record here — what was funded or done, for which moku, what has been reported since — and know exactly where each claim came from.
Open analysis. A researcher, a community group, or a program can take the published records as data: add them up by moku or department, compare across years, join them to other public data, and publish what they find — provided the analysis is theirs and traces back to the record it came from.

The record is the unit of both. What makes it usable for either is the same thing: nothing publishes without passing a person, and nothing published is unsourced.

2. Who reviews what

Three gates stand between an edit to the register and a public page, and a person holds each one:

The record review — a reviewer decides when a record is published. The software checks that a story, at least one result, a source, and the moku are present, but it never publishes on its own. In a report that tracks State funding, the budget facts and the public notices appear for every funded project without this review; the review is for the project's own story and its numbers.
The change review — every edit to the register is recorded and reviewed, and the full history is public. A correction is a change anyone can read.
The release switch — nothing goes live until a person turns publishing on for the site. Until then the checks run and the pages are built, but nothing is released.

The rule all three share: a person publishes, never software.

3. What anyone may do

Cite a record. Any project, its budget, its place as reported, and any entry under What has been reported may be cited as public record, with the record's identifier (GF-2027-HI-082) and the date or version of the report consulted.
Add up and compare. Sums, counts, and shares by moku, island, department, kind of work, or year; comparison across years as reports accrue; joins to other public data (the State's own layers, the Mokunet shared map, the research commons) by the record's identifier, moku, or program ID.
Publish analysis. Indicators, rankings, maps, and models built on the data download, published under the analyst's own name with the method stated and the source records named.
Reproduce the gates. Run the same checks and build on the same register, or on a register kept elsewhere, with the open toolkit the reports are built on.
Review and adapt the method — offices and programs that own real program records may do so under their own authority and review.

4. What no one may do

Represent a commons report as a government system or a State report, or imply agency adoption, endorsement, or partnership. Budgets are the Legislature's; everything reported about them here is recorded from a named public source or from the people doing the work.
Use a commons report — or any real system built on this pattern — to make or justify funding, eligibility, zoning, staffing, enforcement, or permitting decisions. A view that flags items for a closer look is for further professional review; it decides nothing.
Publish derived, extrapolated, or modeled figures as if they were reported results. A derivation is the analyst's and must say so; a reported result is the project's and carries its source note.
Fill in what the record marks to be reported — a place, a moku, an Act 96 purpose — and present it as the record's. Where a report reads a place from a project's name, it says so and asks for confirmation; an analysis that goes further must say the same.
Treat the made-up FY2026 practice year, kept for testing and marked practice data wherever it builds, as real.

5. What the record never stores, and what it holds back

In a report that tracks enacted funding, every funded project appears — on its page and in the data download alike — with the facts of the law and its sourced notices. The project's story and its reported results appear only once a person has published the record; until then the page says so, and the download leaves them empty.
The name and email of whoever submits a report are kept for accountability, never published: removed from every download, shown on no page.
The register holds no other personal information, no financial system of record, and no pre-decisional program material, and it must never be extended to. Parcel numbers are references only; boundaries, ownership, and taxation stay with the county and State parcel maps.
A story is in the voice of the people doing the work, released with their record; field observations are combined before they publish.
Where a report tracks State funding, the State's records are the authority for every fact of the budget; a disagreement is a defect in the register by definition, corrected in a reviewed change.

6. Versions and release readiness

The rules, the toolkit documents, and the reports carry explicit versions; release notes accompany every versioned release. The register is checked against rule set v0.3: a practice report says so on every page it builds and a real one never can; island, moku, and Act 96 purpose may stay blank until reported; a record cannot publish without its moku. A second rule set, v1.0, for a program that tracks budget releases and spending and the people accountable for each submission, is published alongside and tested. The same checks run against a register kept elsewhere. Document 2 gives the field-by-field detail.

Release checklist: rules and checks pass on every change · the review gate exercised (unpublished stories and results verified absent from pages and downloads) · the source chain and the sources under What has been reported present on every page · the public-record notice present on every page and download · accessibility checks passed · release note written.

7. Accessibility

The reports target WCAG 2.1 AA, consistent with the State of Hawaiʻi's EIT accessibility standards: semantic landmarks and headings, skip navigation, keyboard-visible focus, AA contrast, table captions and scoped headers, and no information conveyed by color alone. Accessibility findings are release blockers, not backlog items.

8. Open analysis on the published record

The reports, the register, its rules, and the tools that build the pages are published in the open, so the accountability is itself accountable. For analysis, use the record, not the page:

The data download for each report (registry/FY2027.published.json for the Green Fee report): one entry per record, holding exactly what the pages show. Read from it rather than scraping pages; later reports publish beside it.
Identity is stable. A record's identifier is never reissued or renamed, and its page address never moves once built. Cite the identifier, and the version or date of the download you read.
Every record names its sources. The budget worksheet page and program ID; each notice's source and link; each result's source note; the change that published it. Carry them forward: an analysis that cites a figure cites its source note.
Moku context follows the Mokunet shared map, the 33 moku of Hawaiʻi. It is regional context — whose place — never a land-use classification and never a jurisdiction; analyses by moku should say the same.
Corrections are welcome and public: open an issue on the repository naming the record and the source that contradicts it; a correction is a reviewed change, and the history keeps both versions.
Reproduce the gates with the open toolkit the reports are built on, which runs the same checks and build on any register and proves, on every change, that it still reproduces this one exactly.