@letterstory/cli 0.58.0 → 0.60.0

package/README.md

@@ -360,8 +360,9 @@ letterstory admin tuning comparison <id>               # per-measure verdicts (b
 letterstory admin tuning picks <id>                    # the blind queue; then: admin tuning pick <id> --cell <key> --pick left|right|tie
 letterstory admin tuning reading-list <id> --n 10      # your next blind pairs as links (Text A / Text B), to read off the screen; pick in the console
 letterstory admin tuning pair <id> --cell <key>        # one pair named: live and candidate, both texts, the measures per arm (a key reads every pair)
+letterstory admin tuning side-by-side <id>,<id>,<id> [--cell <key>]   # several writers as one board: each one's measures against live; with --cell, every writer's text on that input
 letterstory admin tuning candidates --kernels all --patch '{"writer":"echo/echo"}' --note "Echo"   # the same change as a candidate of each kernel
-letterstory admin tuning session-new --title "Echo round 1" --kernel Hazel --input-file brief.md --kind outline [--version 22]
+letterstory admin tuning session-new --title "Echo round 1" --kernel Hazel --input-file brief.md --kind outline [--version 22] [--versions 23,29]   # --versions: saved candidates, a trial each
 letterstory admin tuning propose <session> --text "Echo with rarity on; live with the shred after" --create   # trials from words, over the dials
 letterstory admin tuning trial-run <trial> --repeats 2  # runs advance on the server; poll `admin tuning session <id> --texts`
 letterstory admin tuning trial-save <trial> --note "…"  # a trial as a candidate version (never live); then `admin tuning compare`

package/lib/cli.mjs

@@ -55,7 +55,7 @@ import {
 } from "./commands.mjs";
 
 // Keep in sync with cli/package.json.
-export const VERSION = "0.58.0";
+export const VERSION = "0.60.0";
 
 // Flags that never take a value. Listing them explicitly means `deploy get --json <id>`
 // can't accidentally swallow the id as --json's value.

package/lib/commands/admin.mjs

@@ -430,6 +430,7 @@ async function cmdAdminTuningContext(ctx) {
  *   admin tuning pick <comparison-id> --cell <key> --pick left|right|tie   record a blind pick
  *   admin tuning reading-list <comparison-id> [--n 10]          your next blind pairs as links, to read off the screen
  *   admin tuning pair <comparison-id> --cell <key>             one pair named: live and candidate, both texts, the measures per arm
+ *   admin tuning side-by-side <id>,<id>[,…] [--cell <key>]     several comparisons as one board: each writer's measures against live, and one input's texts
  *   admin tuning candidates --kernels Hazel,Aspen|all --patch '{…}' [--base N] [--note "…"]   the same changes as a candidate of each kernel
  *   admin tuning promote <experiment-id> --versions <service-kernel-id>:<version>,…   promote from a "ship" decision
  *
@@ -640,6 +641,34 @@ async function cmdAdminTuning(ctx) {
 				await staffRequest(ctx, "GET", `${T}/comparisons/${id}/pair?${new URLSearchParams({ cell })}`)
 			);
 		}
+		case "side-by-side": {
+			const ids = (id ?? "")
+				.split(",")
+				.map((x) => x.trim())
+				.filter(Boolean);
+			if (!ids?.length) throw new CliError("Usage: admin tuning side-by-side <id>,<id>[,…] [--cell <key>]");
+			const cell = flagStr(flags.cell);
+			const r = await staffRequest(
+				ctx,
+				"GET",
+				`${T}/comparisons/side-by-side?${new URLSearchParams(compact({ ids: ids.join(","), cell }))}`
+			);
+			if (json) return printResult(io, flags, r);
+			for (const w of r.writers) {
+				const m = (measure) => w.measures.find((x) => x.measure === measure);
+				const q = m("quality");
+				io.log(
+					`${w.label} (${w.writers.join(", ") || "—"}, ${w.settled_pairs} pairs, ${w.status}): quality ${q?.mean_delta ?? "—"} [${q?.ci95?.join(" … ") ?? "—"}] ${q?.verdict ?? ""} · seconds ${m("latency_s")?.mean_candidate ?? "—"} vs ${m("latency_s")?.mean_control ?? "—"} · picks ${w.picks.candidate}–${w.picks.control}`
+				);
+			}
+			if (r.input) {
+				for (const t of [r.input.live, ...r.input.texts].filter(Boolean))
+					io.log(
+						`\n=== ${t.label} ${t.writer ?? ""}${t.version ? ` v${t.version}` : ""} (${t.access})\n${t.text ?? "(no text)"}`
+					);
+			} else io.log(`\ninputs: ${r.inputs.map((i) => i.key).join(", ")}  (add --cell <key> for the texts)`);
+			return;
+		}
 		case "candidates": {
 			const patch = jsonFlag(flags.patch, "--patch");
 			const wanted = flagList(flags.kernels);
@@ -730,7 +759,7 @@ async function cmdAdminTuning(ctx) {
 			}
 			if (!title || !kernel || !input)
 				throw new CliError(
-					'Usage: admin tuning session-new --title "…" --kernel <id|name> (--input-file f.md | --input "…") [--kind article|outline|brief] [--version N] [--writers a,b] [--trials-from <session-id>]'
+					'Usage: admin tuning session-new --title "…" --kernel <id|name> (--input-file f.md | --input "…") [--kind article|outline|brief] [--version N] [--writers a,b] [--versions 23,29] [--trials-from <session-id>]'
 				);
 			let kernelId = kernel;
 			if (!/^[0-9a-f-]{36}$/i.test(kernel)) {
@@ -752,6 +781,11 @@ async function cmdAdminTuning(ctx) {
 					base_version: flagNum(flags.version),
 					trials_from_session_id: flagStr(flags["trials-from"]),
 					writers: (flagList(flags.writers) ?? []).map((model) => ({ model })),
+					// Candidates to compare: other versions of the base kernel, a trial each.
+					versions: (flagList(flags.versions) ?? [])
+						.flatMap((v) => v.split(","))
+						.map((v) => Number(v.trim()))
+						.filter((v) => Number.isInteger(v) && v > 0),
 				})
 			);
 			if (json) return printResult(io, flags, { session });
@@ -1141,7 +1175,7 @@ async function cmdAdminTuning(ctx) {
 		}
 		default:
 			throw new CliError(
-				"Usage: admin tuning kernels|kernel|candidate|compare|comparisons|comparison|picks|pick|reading-list|pair|candidates|base-kernels|base-versions|sessions|session|session-new|propose|trial-add|trial-run|trial-save|promote|arenas|arena|arena-candidate|research-compare|research-inputs|research-split|research-standing|promote-arena|shred-compare|shred-inputs|shred-split|shred-standing|shredders|shredder|shredder-candidate|promote-shredder|enhance-compare|enhance-inputs|enhance-split|enhance-standing|enhancers|enhancer|enhancer-candidate|promote-enhancer (see --help)"
+				"Usage: admin tuning kernels|kernel|candidate|compare|comparisons|comparison|picks|pick|reading-list|pair|side-by-side|candidates|base-kernels|base-versions|sessions|session|session-new|propose|trial-add|trial-run|trial-save|promote|arenas|arena|arena-candidate|research-compare|research-inputs|research-split|research-standing|promote-arena|shred-compare|shred-inputs|shred-split|shred-standing|shredders|shredder|shredder-candidate|promote-shredder|enhance-compare|enhance-inputs|enhance-split|enhance-standing|enhancers|enhancer|enhancer-candidate|promote-enhancer (see --help)"
 			);
 	}
 }

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@letterstory/cli",
-	"version": "0.58.0",
+	"version": "0.60.0",
 	"description": "Spin up and manage Letterstory phantom blogs from your terminal.",
 	"type": "module",
 	"license": "UNLICENSED",