feat(toolbox): the judgement through the same door -- validate-proposal, verdict-key, capture-verdict

B's premise applied to the three steps that DECIDE a proposal: a proposal authored outside po --
by a human, or by an agent that is not po -- now meets the blocking deterministic gate, mints the
learning key, and is captured as a Verdict, all without a chat client on the way.

Three thin adapters, no second implementation. The reason is the one the first four doors were
built on, but it bites harder here: the refusal SENTENCE is fed back verbatim into the next
attempt by step 5, so a door that reworded it would break the repair loop while still looking
correct. The probes assert the sentence, not a substring two stages share.

One measurement decided a design detail. The IR writes whole magnitudes as JSON integers
(30000), the run path carries the pydantic float, and verdicts._mint_id hashes the raw value --
so minting from the undeclared JSON would hand out a DIFFERENT verdict id than the debate does
for the same proposal. The door therefore reads the proposal through SavingsProposal and feeds
model_dump() to the public features_from_ir; the probe pins both forms and asserts they differ,
so the shortcut cannot come back silently.

A blocked proposal exits 3, carrying the verdict rather than an exception envelope. "You asked
right and the answer is no" is the same fact whether a file was missing or a claim was
infeasible, and a caller that only reads the exit code must not see a blocked proposal as a
cleared one.

Fasit outside the door in every arm: the base's own checked-in golden suite (written before the
toolbox existed, so it cannot have been fitted to it), a cost baseline authored in the test, a
method cap computed by hand from the fixture, and the public minting rule. Each refusal arm has
an rc-0 control on an argv that would otherwise be accepted.

STATED LIMIT: the input-grounding stage (P7, stage 0b) has no flag here. It falsifies a proposal
against the rendered prompt the model received, and an outside caller has no such prompt.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

README.md

@@ -100,10 +100,24 @@ uv run portfolio-optimiser-toolbox cost-baseline --bundle-dir <base> --project-i
 
 # Admit (or refuse, by name) a declared pre-pass cut before it may shape a run
 uv run portfolio-optimiser-toolbox prepass-admit --payload <cut.json> --bundle-dir <base>
+
+# Run the blocking deterministic gate on a proposal written OUTSIDE the framework.
+# Exit 3 carries the verdict: the verbatim refusal sentence and the stage that wrote it.
+uv run portfolio-optimiser-toolbox validate-proposal \
+  --proposal shared/examples/bygg-energi-mikro/validator-input.json \
+  --cost-baseline <project-prices.json>
+
+# The learning key a verdict on that proposal will arrive under — without deciding anything
+uv run portfolio-optimiser-toolbox verdict-key --proposal <proposal.json>
+
+# Mint an expert's judgement into the Verdict the store holds (stdout is the on-disk form)
+uv run portfolio-optimiser-toolbox capture-verdict --proposal <proposal.json> \
+  --decision approved --rationale "why"
 ```
 
 Each subcommand calls the same function the run path calls — not a copy of it. That is what makes
-the answers you get here the answers the debate gets.
+the answers you get here the answers the debate gets: a proposal a human wrote meets the same
+stages, in the same order, with the same sentence, as one an agent produced.
 
 Verify the install by running the whole suite from the clean clone: