probe: gate grand-exchange-live, the first interactive environment

Eight environments gated, one warning left (tera-crow-nav). The probe calls
measure_ladder.ladder() rather than re-deriving the policies, and seeds from the
skip-aware baskets() helper — a contiguous range() grades a different set of
baskets than the model is served.

Beyond the four-rung ladder it asserts what house rule 4 actually claims: the
turn budget binds in every block of twenty-four, sweeps all 64 cells of the
reference family, scans the rendered turns for held-out ticks, and checks that
hostile replies parse to a hold and score zero.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

README.md

@@ -17,6 +17,7 @@ whatever maxes the string comparison. Nothing here is a port.
 ## The house rules
 
 Every environment in this repository is held to three, and ships the probe that proves it.
+An interactive one is held to a fourth.
 
 1. **Graded on what the model did not see.** If a held-out slice cannot be generated, the
    environment does not measure generalisation and does not ship.
@@ -25,9 +26,15 @@ Every environment in this repository is held to three, and ships the probe that
 3. **A zero floor and a reachable ceiling, demonstrated.** Inaction scores 0.0 and an oracle
    scores 1.0, measured, in `probe.py`. An unreachable component is dead weight in the
    gradient; a floor above zero pays for doing nothing.
+4. **A turn budget that binds.** In an environment the model acts in over several turns, a
+   policy that spends every turn must score measurably below one that spends the turns worth
+   spending. Every reward here saturates, so without a cost for looking the budget is free
+   and the multi-turn form measures nothing the one-shot form did not. In
+   `grand-exchange-live` the gap is 0.369, stdev 0.054 over five blocks of twenty-four,
+   worst block 0.316.
 
-Rule 3 is not decoration. Every environment here was wrong the first time and the probe is
-what caught it — see the table below, and the git history.
+Rules 3 and 4 are not decoration. Every environment here was wrong the first time and the
+probe is what caught it — see the table below, and the git history.
 
 ## The environments
 
@@ -40,6 +47,7 @@ what caught it — see the table below, and the git history.
 | [`bot-detection`](environments/bot_detection) | Classify accounts as bot or human from activity logs, learning the rule from six labelled examples. | Click latency, session length and route count are drawn before the generator decides who is a bot, so every timing rule sits at chance. The bots that cloak leak on one behavioural channel each, and there are three. |
 | [`drop-table-inference`](environments/drop_table_inference) | Estimate a monster's drop rates from a kill log, graded on the next forty thousand kills. | Copying the observed frequencies is right about the common drops and asserts that the item which appeared zero times is impossible. Nothing in the counts distinguishes one silent item from another, so the only evidence about them is the rungs the observed items did not take and what the silent ones are worth. |
 | [`grand-exchange`](environments/grand_exchange) | Read price and volume history for a basket of items and place limit orders, executed against the next thirty ticks. | One or two items per basket are a random walk, not a mean-reverting one — the widest-swinging lines on the board and the ones with no anchor to revert to — and you are not told how many there are, so counting is not a substitute for the shape statistic. And liquidity in gp per tick is uncorrelated with price, so the fattest visible margins sit where the purse cannot go. |
+| [`grand-exchange-live`](environments/grand_exchange_live) | The same basket, traded while the window is running: eight looks over sixty ticks, with stock, offers standing between looks, and a purse that sale proceeds come back to. | Looking is not free. Re-quoting an item that already carries an offer freezes the **whole** book for one to four ticks, an amended offer restarts at the back of the fill queue, and the per-item buy limit is cumulative over the window rather than per offer — so the policy that re-quotes on every look scores 0.631 against a reference that looks four times and scores 1.000. |
 
 ## What the probes measure
 
@@ -54,6 +62,7 @@ Reward for the degenerate strategies and for an oracle, from `uv run python prob
 | `bot-detection` | 0.000 | 0.000 *(ban everyone)* | 0.400 *(one tell of three)* | **1.000** |
 | `drop-table-inference` | 0.000 | 0.179 *(copy the frequencies, floor the rest at a memorised constant)* | 0.476 *(per-item posterior over the grid)* | **1.000** |
 | `grand-exchange` | 0.000 | 0.082 *(buy and sell at market)* | 0.608 *(anchor on the mean and size to the volume, trap included)* | **1.000** |
+| `grand-exchange-live` | 0.000 | 0.126 *(buy and sell at market, on one look)* | 0.406 *(anchor on the mean, no trap filter, one look and no re-quote)* | **1.000** |
 
 ## Running one