@feigi/fleet-ctl 3.144.4 → 3.146.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -247,10 +247,15 @@ and the guard the script also widens a `docs/metrics/member-outcomes.tsv`
247
247
  written before the newest columns: each row gains them, `ending` reads
248
248
  `unknown` on an implementer's row, and nothing else changes. Without that,
249
249
  phase 0, the close-out scrape and `tier-outcomes.mjs append` refuse the file.
250
- After a guard it wrote, the script writes the Preview fit,
251
- `.fleet/router-preview.json`, which only the cockpit reads: nothing routes by
252
- it and nothing commits it, and a `metrics-refresh: preview: failed — <why>`
253
- line leaves the exit at 0, so it never stops the run.
250
+ After a guard it wrote, the script computes the Fit — the router's table fitted
251
+ over the three metrics files and that guard, with the starting table beside the
252
+ scripts as its prior — and writes it to `.fleet/router-fit.json`, which every
253
+ Pull of this run routes by. Nothing commits it. One Fit per run: close-out
254
+ computes none, so each close-out's rows route the next run. A
255
+ `metrics-refresh: fit: failed — <why>` line leaves the previous file as it
256
+ was and the exit at 0, so it never stops the run: the run routes by the last
257
+ Fit that computed, else by the starting table. A skipped guard skips the Fit
258
+ too, and both files are left as they are.
254
259
 
255
260
  ## Launch the cockpit
256
261
 
@@ -276,8 +276,13 @@ one; a B Pull routes exactly as an A one until the table opens arm B.
276
276
  row set of the run's Objective, `OBJECTIVE`, which it reads off the phase-0
277
277
  receipt. The printed line is the
278
278
  whole routing decision — never re-derive `CELL` from the table, the guard or
279
- this paragraph. A missing or unreadable `.fleet/cost-guard.json` routes the
280
- default cell only (`REASON=guard-missing`), burn-in included. On the ledger the
279
+ this paragraph. The table it routes by is the Fit phase 0 wrote,
280
+ `.fleet/router-fit.json` beside `--guard`; a Fit that is absent, unreadable,
281
+ unparseable or not a valid table routes by the starting table beside the
282
+ script, and `REASON` says so (`fit-missing`, `fit-unreadable`, `fit-invalid`).
283
+ A missing or unreadable `.fleet/cost-guard.json` routes the
284
+ default cell only (`REASON=guard-missing`), burn-in included, whatever the Fit
285
+ holds. On the ledger the
281
286
  whole line, `REASON` with it, is the Pull's `## Routed` entry; a row carries no
282
287
  `route=` (an older row's `route=<REASON>` still reads).
283
288
 
@@ -441,7 +446,7 @@ controller's probe could not look.
441
446
  **Every 5th Pull by ledger count goes at an exploration cell.** Count the `impl-`
442
447
  rows in `.fleet/ledger.md` at Pull time; the Pull that creates row 5, 10, 15 …
443
448
  is an Exploration Pull, and the router draws its cell: `k = 1 +
444
- (sha256("<session>\t<ticket>")[0:8] mod K)` over `router-table.json`'s
449
+ (sha256("<session>\t<ticket>")[0:8] mod K)` over the routing table's
445
450
  `cells` minus the policy cell, in token order. Under either Objective that
446
451
  policy cell is the `cost` row set's, so the draw is the same under both. While the table's `burn_in` is
447
452
  true every Pull draws, over every cell, the policy cell included. An