AQ-PUBLIC-CLAIM-GATE — Claim approval procedure and forms
SCOPE: WHAT COUNTS AS A CLAIM
A claim is any statement about Akasha-Q — its architecture, its evidence, its schedule, its relationship to Visvambhara, or its position on the six-rung reach ladder — made to any reader or listener outside the immediate working group. Surface does not change the definition. A slide title, a figure caption, a sentence in a funding memo, a spoken answer in a demonstration, a line in a job description and a caption under a photograph are all claims and all pass the same gate. Internal working notes are not claims; the moment a working note is shown outside the group it becomes one.
The gate decides one question only: may this exact sentence be said, at this evidence level, under this assumption set, on this surface, now. It does not decide whether the underlying engineering is a good idea, and it does not restrict what may be built. It restricts only what may be asserted.
THE SIX STAGES
| Stage | Question answered | Held by | Failure route |
|---|---|---|---|
| S1 Submission | Is there a complete, single-sentence claim with an author who stands behind it? | Claim author | — |
| S2 Admissibility | Is the form complete, are the ids real, is the established vocabulary used without parallel invented terms? | Register custodian (non-voting) | Returned to S1. Not a judgement on truth. |
| S3 Trace audit | Can a non-author walk the three navigation steps to a primary record, unaided? | Navigator, must not be an author of any artefact in the trace | Held. If the claim is already posted, it is removed from the Evidence Wall immediately. A blocked path cannot be overridden by the panel. |
| S4 Technical review | Is the sentence entailed by the trace at the stated evidence level under the stated assumption set — not merely consistent with it? | Seated panel | Defer, or revise after more evidence. |
| S5 Wording adjudication | Does the sentence survive the constrained-language rules, the prohibited-claim list, the numeric-invention rule and the exclusion of geopolitical, sovereign-capability and defence framing? | Seated panel, chaired | Reject the sentence. A different sentence may be submitted the same day. |
| S6 Disposition | What is the recorded decision, the bound text, the qualifier if any, the surface and the expiry? | Chair and custodian | — |
Figure 1 - stage flow and the five terminal states
+---------------------------------------------------------+
| S1 SUBMISSION author files form fields F-01..F-16 |
+---------------------------------------------------------+
|
v
+---------------------------------------------------------+
| S2 ADMISSIBILITY complete? ids real? vocabulary exact? |
+---------------------------------------------------------+
| pass | incomplete --> back to S1
v
+---------------------------------------------------------+
| S3 TRACE AUDIT non-author walks the 3 steps |
+---------------------------------------------------------+
| navigable | blocked --> HOLD, and
| | REMOVE from wall if posted
v
+---------------------------------------------------------+
| S4 TECHNICAL entailed at the stated level? |
+---------------------------------------------------------+
|
v
+---------------------------------------------------------+
| S5 WORDING constrained-language adjudication |
+---------------------------------------------------------+
|
v
+---------------------------------------------------------+
| S6 DISPOSITION one state, recorded in L6 |
+---------------------------------------------------------+
|
+--> approve -> wall entry created
+--> approve with qualification -> qualifier bound into text
+--> defer -> returns to a named stage
+--> revise after more evidence -> named experiment or level
+--> reject -> sentence barred, recorded
THE FIVE DECISION STATES
The gate deliberately reuses the decision grammar of L5 command authority. Q1 who submitted this, Q2 is the trace valid, Q3 is this surface authorised for this claim now, Q4 should it be said at all given the current evidence state. Q1–Q3 are hard constraints; Q4 can only reduce what Q1–Q3 allowed, never enlarge it.
| State | Analogue | Effect | Exit condition |
|---|---|---|---|
| Approve | execute | The exact sentence may be used on the named surfaces until expiry. It is posted to the Evidence Wall with its seven fields. | Expiry, level fall, or strike. |
| Approve with qualification | execute-restricted | Approved only as the sentence plus a bound qualifier clause — assumption set, evidence level, or an explicit "no target is asserted". The qualifier is part of the sentence and may not be moved to a footnote, a following slide, or a verbal aside. | As above; dropping the qualifier voids the approval. |
| Defer | safe-hold | The panel cannot decide now for procedural reasons: no independent navigator available, quorum unmet, or an unresolved conflict CF-1..CF-7 touching the claim. Nothing may be said in the interim. | Returns to the named stage with the blocking condition recorded. |
| Revise after more evidence | abstain | The sentence is not entailed at the level claimed. The panel must name either the lower evidence level at which it would be entailed, or the specific artefact that would entail it. "Insufficient confidence to authorize" is a permitted and expected outcome (DC-3). | Resubmission at the lower level, or after the named artefact exists. |
| Reject | deny | The sentence may not be used on any surface. It is recorded in the rejected-sentence list so that later paraphrases of it can be detected at admissibility. | None for that sentence. A materially different sentence is a new submission. |
Any sentence asserting that the system is unhackable, impossible to intercept, universally quantum-secure, information-theoretically secure at system level, satellite-ready, space-proven or operational. Any sentence using secure, resilient, autonomous, high-precision, low-latency or quantum-secure as an unqualified descriptor rather than naming the measured quantity, the assumption set and the evidence level. Any numeric performance target not traceable to a measurement record. Any geopolitical, sovereign-capability, national-security or defence framing. Any statement placing Akasha-Q as a subsystem of Visvambhara, or presenting Visvambhara as the reason Akasha-Q exists. Any statement making external capital, partnership or government collaboration a precondition of the programme continuing.
REVIEW COMPOSITION AND INDEPENDENCE
- Quorum: three seated reviewers plus the custodian, who convenes, records and does not vote.
- Standing seat: L9 systems engineering & safety governance holds one seat at every sitting.
- Independence requirement: at least one seated reviewer must have authored no artefact in the claim's trace and must not sit in the claim author's reporting line. If no such person is available, the only lawful outcome is defer — the sitting does not proceed on the strength of goodwill.
- Navigator separation: the S3 navigator is never an author of the trace and never the chair. The navigator's finding on navigability is binding and is not a vote.
- Recusal: the claim author may attend, must answer questions, and may not vote or chair.
- L8 is advisory only: AI-assisted analysis may submit written notes pointing reviewers at artefacts or at internal inconsistencies. It holds no seat, casts no vote, and may never be cited as a supporting artefact for the truth of a claim. Its notes are recorded as advisory in the disposition.
- Blocking: a single reject at S5 blocks approval of that sentence; approval at S4 requires no dissent on entailment, with any dissent recorded verbatim in the disposition.
CLAIM SUBMISSION FORM
| Field | Content required |
|---|---|
| F-01 | Submission id — assigned by the custodian on receipt, not by the author. |
| F-02 | Claim id requested: new, or a revision of an existing AQ-CL-nnnn. |
| F-03 | The exact claim sentence. One sentence. The text approved is the text used. |
| F-04 | Claimed evidence level — exactly one rung of the Part 12 ladder. |
| F-05 | Architecture tier: CA, Y1, FR or LH. |
| F-06 | Layers L0–L9 and trust boundaries TB-1..TB-5 the claim touches. |
| F-07 | Assumption set id, plus its assumptions written out in plain sentences. |
| F-08 | DC-1 tagging: every quantity named in the sentence marked MEASURED, DERIVED or MODELLED. |
| F-09 | Supporting artefacts: ids, types, dates, authors, and the assumption set each was produced under. |
| F-10 | Trace reference AQ-EB-nnnn-x and the three-step path as the author believes a stranger would walk it. |
| F-11 | Replication status: none / internal re-run / independent internal / external, naming the replicator. |
| F-12 | Limitations, and the specific observation that would falsify the claim. |
| F-13 | DC-2 declaration: quantities relevant to the sentence that remain undetermined, listed explicitly rather than omitted. |
| F-14 | Numeric declaration: every number in the sentence with its origin record, or the statement "no numeric target is asserted". |
| F-15 | Intended surfaces and the demonstration maturity DM-0..DM-6 at which the sentence would be used. |
| F-16 | Author, co-authors, declared conflicts, requested validity window, and any prior claim this supersedes or strikes. |
PARAPHRASE IS AN UNAPPROVED CLAIM
Approval attaches to a sentence, not to a subject. Any restatement of an approved claim — a shortened slide title, a translation, a spoken answer that compresses the qualifier, a caption, a summary in an email — is a new and unapproved claim, and using it is a gate violation regardless of whether the speaker believes it means the same thing. The reason is mechanical: the qualifier and the evidence level are the parts a paraphrase drops first, and dropping them converts a bounded statement into a prohibited one.
Operationally, every person who speaks for the programme carries the approved-sentence card for the surface they are on: the exact approved sentences, their chips, and their bound qualifiers. Off-card questions have one approved answer pattern — name the metric, state that it is undetermined, and state how a target will be established from measurement. Refusal is a permitted output, not a failure of the speaker (DC-3). After every sitting and every demonstration, off-card questions are logged and enter the gate at S1 as candidate claims.
Every disposition — including reject and defer — is written to the L6 Evidence & Provenance Fabric with the submission form, the panel roster, the independence finding, dissent, and the bound text (DC-5). Approvals expire at the next programme review unless renewed, and lapse immediately on any level fall against a claim in their trace: degradation does not wait for a meeting, and restoration runs the full sequence (DC-4).