Hygiene runtime is standard-rich/runtime-poor: no CTEST runner, no passport producer, LOGFALL id-namespace drift

[mdheller]

Seated under ADR-0004 (the Crown) — this is the runtime that will produce the Bias/Calibration Passports the constitution's Truth Engine layer references.

The epistemic-governance Hygiene standard (sociosphere standards/epistemic-governance/, ruleset 1.3.0) is standard-rich but runtime-poor:

1. No CTEST runner. ADR-0004 invariant T1 requires every admitted Claim/TruthRecord to carry a TestObligation (refutation channel). The refutation channels are specified but nothing runs them.
2. No bias-passport / calibration-passport producer. Spec 14 names the Calibration/Bias Passport as a per-participant reasoning credential; there is no runtime that emits one.
3. Id-namespace drift. Noetica emits LOGFALL.ADHOMINEM.V1 at ruleset 0.1.0 while the standard defines LOGFALL.ADHOM.V2 at 1.3.0 — the detector output and the standard's vocabulary disagree on both the id and the version.

Action: build the CTEST runner + passport producer against ruleset 1.3.0, and reconcile the LOGFALL id namespace (single canonical id + version) so detectors and standard agree.

Ref: ADR-0004 docs/adr/0004-crown-telos-truth-constitution.md; spec 14 docs/specs/14_Truth_Engine_Discourse_Integration_v0_1.md.

[mdheller]

P0 Hygiene runtime — landed as 4 consume-not-fork PRs (seated under the Crown, ADR-0004 / #52)

The standard-rich / runtime-poor Hygiene layer is now a live, teeth-verified runtime. One PR per repo touched:

Deliverable	PR	Teeth
Id-namespace reconciliation (governed detector-id map) + set-1 bias catalog	SocioProphet/sociosphere#554	bijection; every standard_id exists + is shipped-maturity; round-trip both ways; --selftest rejects drift / V1→V2 laundering / incomplete allow-list
CTEST runner + Bias/Calibration Passport producers (the runtime)	SocioProphet/agentplane#331	no falsifier → epistemically_void (T1); affirming-the-consequent → REJECTED_AFFIRMING_THE_CONSEQUENT (D1); passing counter-test → admitted; ungoverned detection → rejected; non-conformant output → rejected; sealed receipts; 15/15 selftest
Detector-id boundary — a Noetica firing emits the STANDARD id	SocioProphet/Noetica#609	every shipped detector reconciles to a governed id; drift rejected; tsc clean
Ontogenesis concept registration (bias-passport, calibration-passport, ctest-run, hygiene-profile, detector-id-map)	SocioProphet/ontogenesis#142	5/5 admitted with source anchors + provenance + tamper-evident receipts

Seat under the Crown (bind-upward)

• T1 (no Test-Obligation → void) ← run_ctest() epistemically_void, mirroring the SILENT Phase-0 gate (evidence-intake-kernel#2).
• D1 (Da'at cannot assert truth = affirming-the-consequent guard, eik#3) ← LOGFALL.AFFIRMCONSEQ.V1 + the runner's REJECTED_AFFIRMING_THE_CONSEQUENT — the same rule, cross-referenced in bias-catalog.yaml and docs/epigov-hygiene-runtime.md.
• Passports feed Truth Records; CTEST dispatch reuses Noetica reasoner.ts CTEST_ROUTING / PR #570 by reference.

Deliberate calls (flagged for review)

• No V1→V2 alias. The standard advanced to ruleset 1.4.0, which already adopted the shipped V1 ids alongside proposed V2s. The map reconciles emitted→shipped standard id (identity today) + records succeeds_into; it does not emit a V2 id for a V1 regex hit (that would launder a not-yet-built detector). This is a considered deviation from the issue's literal LOGFALL.ADHOMINEM.V1 → LOGFALL.ADHOM.V2 example.
• Ruleset bumped 1.4.0 → 1.5.0 (additive proposed reservations for the set-1 biases + LOGFALL.AFFIRMCONSEQ.V1).

Follow-ups for @mdheller

1. Build runners for the newly-reserved proposed CTESTs (CTEST.FORESIGHT-BLIND.V1, CTEST.REFRAME-INVARIANCE.V1, CTEST.FORMAL-VALIDITY.V1) so the set-1 bias warn/block stop downgrading.
2. Wire the sealed passports into the Truth Record intake on the Truth Engine side.
3. Record this seat in ADR-0004's follow-up section once the PRs merge (left to you — the constitution is board-governed).