the vertical slice's four local repairs

agent-decided, reversible. Taken by prompts/2026-08-25-end-to-end-vertical-slice/, whose report states each with its reason. Reversal is by re-ruling.

**The operative waist is PForce, not a new PInjunction.** The dispatch proposed a new standing payload. Reflective Integrity already carries PForce (commitRef, schemaRef, compiledClause : Clause), Clause is already opaque, and §35's O_t is already its projection; adding a second constructor would have duplicated a payload and made the projection read two where it reads one. Rejected alternative: PInjunction as specified, which would have been honest to the dispatch and redundant in the record.

**The value waist is a new PValue payload constructor.** Licensed by Meta-Stability's conservative-extension rule and free in practice, since delta's three clauses write a payload into a fresh standing state without inspecting it. Rejected alternative: carrying a value specification as PCmt (NonStanceBearing, code), which would have forced a meta-level projection to inspect object-level content and violated §3's stratification.

**The vertical-slice projection is O_n = {(i, J_i)}.** §35's projection is a set of clauses; enforcement provenance and per-term conflict attribution both need the standing identity, and two active standings with equal payloads collapse in a set. No store, constructor or conservation law changes, so this is a richer read of the same fold rather than a reopening. Rejected alternative: carrying provenance inside the clause, which §5 of the dispatch prohibits and which would have made the payload non-frozen.

**sem_L is a third parametric interpreter.** Alongside [[.]]_S and [[.]]_D, keyed by settlement id, with rigidity, finiteness and computability as its assumptions. Rejected alternative: a content field on Settlement, which would have changed a record type in the frozen core's §33 signature to buy the same thing.