VERIFY · LIVE
The tally
Every figure this estate measures about itself, split by who has to act for it to move.
Figures somebody else has to move
1 of 4Figures we can raise ourselves
12Almost every number an organisation can measure about itself is a number it can also raise. Entities, relations, adoption of its own standards, commits sealed — all of them go up when a property talks more about itself, and a reader has no way to separate them from the few that only move when somebody else acts. This page keeps both halves visible, in that order. The split is not editorial: it is a required field on every figure, and a tally/1 document carrying no figure on the outside half does not validate.
Where a figure below carries n of them ours, that part of the numerator is an organisation this estate itself defines — a sibling agreeing with us, which is corroboration of a much weaker kind. A figure whose numerator is entirely ours is refused by the format rather than printed.
Somebody else has to act
1 of 4 figures that require an outside act reads above zero.
| Figure | Reading | Measured by |
|---|---|---|
| Claims about outside organisations they have signedThe denominator counts only claims whose far end names an identity that could sign. A denominator we can pad by pushing is the one thing this figure must not have. | 0 of 30 | tools/estate-countersigned.mjs |
| Organisations outside this estate publishing any of these formatsCaptive adoption — a client, a portfolio company, a shareholder’s business — is recorded separately and never added here. A company whose investor asks it to publish a file publishes the file; that is a fact about the relationship. | 0since 2026-08-01 | adoption-signal.json |
| Claims the counterparty has announced in their own wordsAnnouncing the same fact is not signing our record. Counted here because it is still somebody else’s act, and never counted as a countersignature. self_attributed is the part where the counterparty is an organisation this estate defines — a sibling agreeing with us. | 4 of 303 of them ours | tools/estate-countersigned.mjs |
| Observances a second party has witnessedUnwitnessed practice moves nothing. The workflow that performs an observance can never witness it. | 0 of 2 | tools/estate-ritual.mjs |
We move these ourselves
Worth measuring, worth failing a build over, and worth nobody’s trust on its own. Every one ships with its denominator; a count ships with the date it counts from.
| Figure | Reading | Measured by |
|---|---|---|
| Filed intentions a stranger can readfile() writes private and only a named human promotes. A zero here is the consent rule working, not a broken emitter. | 0 of 65 | tools/estate-intent.mjs |
| Charter served — repositories holding it | 20 of 21 | tools/estate-standards.mjs |
| Checkpoint — repositories holding it | 33 of 33 | tools/estate-standards.mjs |
| Observances a person has consecratedA consecration names a person, and the person is ours. Consecrating our own practice is not evidence about the practice — it sits here for that reason, beside the witness count that does not. | 0 of 2 | tools/estate-ritual.mjs |
| Entities named by more than one property | 25 of 3881 | tools/estate-graph.mjs |
| Decaying claims still current | 677 of 677 | tools/estate-graph.mjs |
| Observances performed across the estateAn observance is performed by an agent on a rhythm. Performing more of them is entirely within our power, which is why it sits on this side and why the rung below is the one that means anything. | 2since 2026-09-03 | tools/estate-ritual.mjs |
| Published relations naming an asserter the graph definesRead 9% in August. A citation to an identity nothing declares is a citation to nothing. | 7368 of 7368 | tools/estate-graph.mjs |
| Sealed entries a stranger can read | 3169 of 3274 | tools/estate-tally.mjs |
| Record-level relations that cross a property boundaryThe honest test of “one graph”. Every other figure here rises when a property talks more about itself; this one needs a boundary crossed. | 143 of 585 | tools/estate-graph.mjs |
| Ritual — repositories holding it | 2 of 42 | tools/estate-standards.mjs |
| Sealed shipped/1 entries across the estateEach entry is one first-parent commit, hashed and append-only. It cannot be edited after the fact; it can be added to by pushing, which is why it sits on this side of the page. | 3274since 2022-12-25 | tools/estate-tally.mjs |
Not measured
Reported rather than dropped or written as zero. An invented zero and a real zero look identical to every reader and mean opposite things.
| Figure | Why not | |
|---|---|---|
| metaverse-ventures — not read this run | the checkout could not be read on the machine this ran on | |
The seal
Each figure carries a sha256 over its own canonical form; the document carries one over the ordered list of those and over every retirement, so the commitment covers what is no longer reported as well as what is. Sort every key at every level, keep array order, hash as UTF-8, and you get the same string.
- As at
- 2026-09-04
- Subject
- org/flashyos
- Digest
4cb9e461246d48ecf75374adf4beea7ccb3df19db71a925b3d72fbef11c24750
Checking it yourself
Check it yourself against /.well-known/tally.json: sort every key at every level, keep array order, hash as UTF-8. It is the same canonicalisation the settlements on /verify/ use, and the vendored implementation in this repository is byte-identical to the published one — a test fails if it drifts. What sealing proves is narrow and real: no figure was edited after the fact. It does not make this tamper-evident. For that you need a witness who is not us, which is exactly what the first table is counting.
What this is not
Not a trend, not a target, not a ranking. A reading beside a goal is read as progress toward the goal, which is a claim about the future; a ranking of properties against each other manufactures the behaviour the estate refuses one layer up. Superlatives and currency figures are refused by name — a superlative is a claim with no denominator, and a money column would be the only column anybody read. The countersignature page is the long form of the first row here.