@azure-id/orc 1.9.0 → 1.9.2

package/bin/cli.js

@@ -180,8 +180,10 @@ function writeStdoutSync(str) {
 }
 
 // Print the object and (optionally) exit with the human path's code.
-function emitJson(obj, exitCode) {
-  writeStdoutSync(JSON.stringify(obj, null, 2) + "\n");
+// `compact` (v1.9.1 B1) writes the answer with no indentation: a `--brief`
+// answer is read by a model, and the card inside it carries its own newlines.
+function emitJson(obj, exitCode, compact) {
+  writeStdoutSync((compact ? JSON.stringify(obj) : JSON.stringify(obj, null, 2)) + "\n");
   if (exitCode !== undefined) process.exit(exitCode);
 }
 
@@ -1065,6 +1067,8 @@ function upgrade() {
 // ---------------------------------------------------------------------------
 
 const KNOWN_MODELS = [
+  "claude-opus-5-5",
+  // Kept so a config written before v1.9.2 still validates.
   "claude-opus-5",
   "claude-opus-4-8",
   "claude-opus-4-7",
@@ -1251,7 +1255,7 @@ const CONFIG_FAMILIES = {
         // without being a CONFIG_META key, because where it sits in the
         // precedence is a fact even though `orc config set` refuses to write it.
         registry_less: true,
-        shadow_note: "shadowed by {by} — executors use the fixed 2-band Opus 5 ladder",
+        shadow_note: "shadowed by {by} — executors use the fixed 2-band Opus 5.5 ladder",
       },
       { prio: "P3", key: null, terminal: "the shipped score→model table" },
     ],
@@ -1438,7 +1442,7 @@ const CONFIG_META = [
   // decision with a recorded reason (the /orc-pact retirement rule).
   { key: "extra_demote_after", def: 2, tier: "common", answers: [{ family: "extra", prio: "P2", mode: "replace" }], gated_by: "extra_enabled", lanes: [], validate: vInt(0), options: [0, 2, 3, 5], desc: "Consecutive `stalled` dispatches on ONE profile, inside one run, before that profile is DEMOTED to the bottom of the ladder for the rest of the run — so `opus5_only` (or the shipped score table) becomes the effective P0 and the work stays on Claude. Only `stalled` counts: a 401 or a rate limit has its own answer (`extra_on_failure`, the vault, `extra_resume`), and demoting on one would hide a credential problem behind a routing change. A resume of the same stalled attempt is the SAME stall, never a second one. A demotion is RUN state — it never writes your config — it is ANNOUNCED before the next dispatch, and it is never auto-promoted back: `orc extra promote <run> --reason \"<why>\"`. 0 turns this clock off and leaves only extra_demote_stale_min." },
   { key: "extra_demote_stale_min", def: 20, tier: "common", answers: [{ family: "extra", prio: "P2", mode: "replace" }], gated_by: "extra_enabled", lanes: [], validate: vInt(0), options: [0, 10, 20, 45], desc: "Minutes a LIVE foreign attempt may show no observable progress before its profile is demoted. This is a different question from `extra_stall_s`, which stops ONE dispatch after 180s of silence: this clock is about the RUN — two workers in flight, both quiet, and a wave that is going nowhere — so the two have their own budgets and their own off values and neither is a simplification of the other. It reads the journal's own progress file on disk, never a remembered fact. 0 turns this clock off and leaves only extra_demote_after." },
-  { key: "opus5_only", def: false, tier: "common", answers: [{ family: "executor-band", prio: "P1", mode: "replace" }, { family: "fixed-role-model", prio: "P1", mode: "replace" }], lanes: ["orc", "orc-analyze", "orc-challenge", "orc-claude", "orc-diy", "orc-doc", "orc-fast", "orc-mini", "orc-pattern", "orc-quick", "orc-retro", "orc-wiki"], validate: vEnum("true", "false"), options: ["true", "false"], desc: "EVERY dispatched role uses ONE model — Opus 5 — with EFFORT as the cost dial (executors: [0,40) low · [40,80) medium · [80,100] high; each fixed role its own pinned effort). Deep SWE-benchmark work on cost vs efficiency across Claude models finds a single Opus 5 agent with the effort ladder the most efficient setup. It FORCES: while on it outranks a hand-written rubric_bands_override. Needs an Opus 5 main session or EVERY dispatch silently downgrades. Excludes the Haiku trace writer and orc-diy (compile-owned)." },
+  { key: "opus5_only", def: false, tier: "common", answers: [{ family: "executor-band", prio: "P1", mode: "replace" }, { family: "fixed-role-model", prio: "P1", mode: "replace" }], lanes: ["orc", "orc-analyze", "orc-challenge", "orc-claude", "orc-diy", "orc-doc", "orc-fast", "orc-mini", "orc-pattern", "orc-quick", "orc-retro", "orc-wiki"], validate: vEnum("true", "false"), options: ["true", "false"], desc: "EVERY dispatched role uses ONE model — Opus 5.5 — with EFFORT as the cost dial (executors: [0,40) low · [40,80) medium · [80,100] high; each fixed role its own pinned effort). Deep SWE-benchmark work on cost vs efficiency across Claude models finds a single Opus 5.5 agent with the effort ladder the most efficient setup. It FORCES: while on it outranks a hand-written rubric_bands_override. Needs an Opus 5.5 main session or EVERY dispatch silently downgrades. Excludes the Haiku trace writer and orc-diy (compile-owned)." },
   // --- v0.46.0 — the six new lanes ------------------------------------------
   { key: "pact_gate", def: "warn", tier: "common", answers: [{ family: "pact", prio: "P2", mode: "replace" }], lanes: ["orc", "orc-pact"], validate: vEnum("off", "warn"), options: ["off", "warn"], desc: "Invariant ledger at Phase 1 + planning: warn = print the one pact line and inject a DRIFTED/BROKEN promise whose anchors intersect the plan's declared files as a planner constraint; off = nothing. NEVER blocks — a promise is advice with a receipt, not a gate. See /orc-pact." },
   { key: "pact_recheck_on_verify", def: "true", tier: "common", answers: [{ family: "pact", prio: "P2", mode: "replace" }], lanes: ["orc", "orc-pact"], validate: vEnum("true", "false"), options: ["true", "false"], desc: "Phase 6: re-run the cheap checks for ONLY the invariants the change touched (`orc pact check`), so a promise that just leaked is caught in the run that broke it." },
@@ -1521,7 +1525,7 @@ const LEGACY_KEYS = {
 // So `config list` prints them, `config list --json` carries them, and
 // `config set` refuses them BY NAME instead of with a generic unknown-key list.
 const RETIRED_KEYS = {
-  fable5_enabled: { removed_in: "1.0.0", why: "the Fable 5 role override was removed — every role dispatches its shipped Claude agent, or the Opus 5 variant under opus5_only" },
+  fable5_enabled: { removed_in: "1.0.0", why: "the Fable 5 role override was removed — every role dispatches its shipped Claude agent, or the Opus 5.5 variant under opus5_only" },
   fable5_roles: { removed_in: "1.0.0", why: "the Fable 5 role override was removed" },
   fable5_effort: { removed_in: "1.0.0", why: "the Fable 5 role override was removed" },
 };
@@ -2080,7 +2084,7 @@ function shadowReason(key, map, claudeDir) {
 //
 // v1.0.0 W4 — TWO BANDS, not three, and the same 90 edge as the default table's
 // top two rows (D13). That symmetry is the point: once the default table's high
-// end is already Opus 5 with effort as the dial, the forcing mode differs from
+// end is already Opus 5.5 with effort as the dial, the forcing mode differs from
 // it only BELOW 65, so a third band here would be a distinction the default
 // table stopped making.
 //
@@ -2126,7 +2130,7 @@ function scoreTableJson(map, claudeDir) {
   //     > the default 6-band table
   //
   // Extra is an OVERLAY, not a replacement, which is what makes "cheap grunt
-  // work goes to DeepSeek, hard work stays on Opus 5" a two-command setup
+  // work goes to DeepSeek, hard work stays on Opus 5.5" a two-command setup
   // rather than a full table rewrite. `active` can therefore read as a
   // COMPOSITE — because the truth is a composite, and a single word would be
   // a lie about what the next dispatch will do.
@@ -2371,7 +2375,7 @@ function extraLegacyRoleWarn(value) {
 
 // A shadowed setting must never be silent — the v0.36.0 rule, applied to a
 // shadow that runs BOTH WAYS. Turning on opus5_only while route rows exist
-// does not give you the Opus 5 ladder everywhere, and turning on Extra does
+// does not give you the Opus 5.5 ladder everywhere, and turning on Extra does
 // not take the Claude table away; saying either would be wrong. So name the
 // ranges, which is the only statement that is true from both sides.
 function extraShadowNotice(claudeDir) {
@@ -2432,17 +2436,17 @@ function opus5Notice(on, claudeDir) {
   console.log(
     "\n  " + ui.color.bold("Opus-5-only dispatch is now ACTIVE — for EVERY role, not just executors.") + "\n" +
       "\n  Why: deep SWE-benchmark work on cost vs efficiency across Claude models finds a\n" +
-      "  single Opus 5 agent, with the EFFORT ladder as the cost dial, the most efficient\n" +
+      "  single Opus 5.5 agent, with the EFFORT ladder as the cost dial, the most efficient\n" +
       "  configuration. You trade model-class variety for effort variety.\n" +
       "\n  Scored executors (/orc + /orc-ultra):\n\n" +
       ladderTable(OPUS5_SCORE_TABLE) +
       "\n  Fixed roles now dispatched instead of their defaults:\n\n" + roleRows + "\n" +
-      "\n  Already Opus 5, unchanged: analyst · planner · reviewer · verifier · test-author\n" +
+      "\n  Already Opus 5.5, unchanged: analyst · planner · reviewer · verifier · test-author\n" +
       "  · combiner · learn-writer · advisor · judge.\n" +
       "  NEVER forced: orc-trace-writer-haiku-4-5 (it transcribes a packet, no reasoning)\n" +
       "  and orc-diy (its table is compile-owned — re-run `orc diy compile` to change it).\n" +
       "\n  " + ui.mark.warn("Tier requirement — read this one:") + "\n" +
-      "  Today only the [90,100] executor band needs an Opus 5 main session. With this ON,\n" +
+      "  Today only the [90,100] executor band needs an Opus 5.5 main session. With this ON,\n" +
       "  EVERY dispatch does. A subagent can never outrank the main session, so on a lower\n" +
       "  session every role silently falls back to the session model and the tier-honesty\n" +
       "  rule reports a downgrade on EVERY return instead of occasionally. Hooks cannot\n" +
@@ -2858,17 +2862,17 @@ const LANE_INERT = {
     { key: "extra_fallback_agent", reason: "re-opening the gate IS the ask, so a second menu composed from a config key would be the same question twice in different words" },
   ],
   "orc-challenge": [
-    { key: "opus5_only", reason: "every agent in this lane is already claude-opus-5, so this is a no-op — the lane is unaffected, not exempt" },
+    { key: "opus5_only", reason: "every agent in this lane is already claude-opus-5-5, so this is a no-op — the lane is unaffected, not exempt" },
   ],
   "orc-doc": [
-    { key: "opus5_only", reason: "both agents in this lane are already claude-opus-5, so this is a no-op — the lane is unaffected, not exempt" },
+    { key: "opus5_only", reason: "both agents in this lane are already claude-opus-5-5, so this is a no-op — the lane is unaffected, not exempt" },
   ],
-  // v1.5.0 — both agents in this lane ship as claude-opus-5, and the
+  // v1.5.0 — both agents in this lane ship as claude-opus-5-5, and the
   // interpreter's `low` is a MEASUREMENT choice rather than a cost one: a
   // harder-thinking interpreter reasons its way to why a leaked stack trace is
   // probably fine in staging, which is exactly the gap it exists to find.
   "orc-test": [
-    { key: "opus5_only", reason: "both agents in this lane are already claude-opus-5, so this is a no-op — the lane is unaffected, not exempt" },
+    { key: "opus5_only", reason: "both agents in this lane are already claude-opus-5-5, so this is a no-op — the lane is unaffected, not exempt" },
   ],
   // v1.1.0 W2 — this lane DISPATCHES NOTHING. A detached command does the
   // waiting, so every family that answers "which model runs this" has no work
@@ -3143,7 +3147,7 @@ function laneAnnounce(lane, map, claudeDir, families) {
     out.push(
       off
         ? `opus5_only: ON, and INERT in this lane — ${off.reason}`
-        : "opus5_only: ON — every dispatched role uses Opus 5, with effort as the only cost dial"
+        : "opus5_only: ON — every dispatched role uses Opus 5.5, with effort as the only cost dial"
     );
   }
   if (isTrue(map.extra_enabled)) {
@@ -3182,9 +3186,10 @@ function laneAnnounce(lane, map, claudeDir, families) {
   const graphOn = String(Object.prototype.hasOwnProperty.call(map, "code_graph") ? map.code_graph : metaFor("code_graph").def) === "on";
   if (graphOn && LANE_CALLS["graph-status"].lanes.includes(lane))
     out.push(
-      "graph: on — use the code-graph cache: (1) now: `orc graph status --if-enabled --heal --json` (builds or updates it); " +
-        "(2) every code-writing slice: `orc graph ctx <declared files> --if-enabled --json` cards; " +
-        "(3) after every code change: `orc graph update --if-enabled --json` (keeps it for the next run). Copy each `line` and `trace` verbatim"
+      "graph: on — use the code-graph cache: (1) now: `orc graph status --if-enabled --heal --json --brief` (builds or updates it); " +
+        "(2) every code-writing slice: `orc graph ctx <declared files> --if-enabled --json --brief` cards; " +
+        "(3) after every code change: `orc graph update --if-enabled --json --brief` (keeps it for the next run). " +
+        "Copy each `line` and `trace` verbatim; `--brief` drops the rows you never print"
     );
   // A shadowed setting must never be silent, and neither must an inert one.
   const shadowed = [];
@@ -3480,7 +3485,24 @@ const LANE_CALLS = {
     on_absent: "exit 4 is an ANSWER, not a miss — plan exactly as before",
     canonical: "_shared/code-graph.md",
     never: "never read a co-change as a DEPENDENCY — it says what people changed together, never what needs what",
-    lanes: ["orc", "orc-diy", "orc-mini"],
+    // v1.9.1: mini no longer calls it. `impact --complexity` answers the same
+    // question inside the SAME process, against the same per-HEAD cache.
+    lanes: ["orc", "orc-diy"],
+  },
+  // v1.9.1 A2 — a USER command. `lanes: []` is the contract: nothing a lane
+  // does is made better by an audit, and a lane that ran one would pay for an
+  // answer about the PARSER in a context that is about the code.
+  "graph-audit": {
+    cmd: "orc graph audit [--top=N] --if-enabled [--json]",
+    what: "WHY the map is thin — density by language, which extractor read each one, the files the parser read as EMPTY longest first, and what their first declaration looks like",
+    exits: { 0: "answered", 1: "no index yet", 3: "off — `code_graph` is off" },
+    states: null,
+    cost: "free",
+    when: "by hand, after `orc graph status` says THIN — never in a lane",
+    on_absent: "exit 1 means there is no graph to audit; build one first",
+    canonical: "_shared/code-graph.md",
+    never: "never read an empty file as a defect — a component with no named function is a file the parser read correctly; only a shape marked `parser gap` is worth an issue",
+    lanes: [],
   },
   "graph-gain": {
     cmd: "orc graph gain --run <trace name> --if-enabled [--json]",
@@ -4962,7 +4984,7 @@ const sha256 = (s) => crypto.createHash("sha256").update(s).digest("hex");
 // The tier-clip roster. It must contain EVERY agent the default table names —
 // `diyScoreTable()` looks each row's agent up here, so a table row naming an
 // agent this map does not know is a crash, not a fallback. v1.0.0 W4 added the
-// two Opus 5 rows for exactly that reason.
+// two Opus 5.5 rows for exactly that reason.
 //
 // It also feeds `fixed_executor`'s option list, so every entry here is offerable
 // — which is what keeps the four agents no band names reachable (D14).
@@ -4994,10 +5016,10 @@ const DIY_TIERS = {
   "opus-4-8-high": { model: 5, effort: 2, modelId: "claude-opus-4-8", effortName: "high" },
   "opus-4-8-xhigh": { model: 5, effort: 3, modelId: "claude-opus-4-8", effortName: "xhigh" },
   "opus-4-8-max": { model: 5, effort: 4, modelId: "claude-opus-4-8", effortName: "max" },
-  "opus-5-med": { model: 6, effort: 1, modelId: "claude-opus-5", effortName: "medium" },
-  "opus-5-high": { model: 6, effort: 2, modelId: "claude-opus-5", effortName: "high" },
-  "opus-5-xhigh": { model: 6, effort: 3, modelId: "claude-opus-5", effortName: "xhigh" },
-  "opus-5-max": { model: 6, effort: 4, modelId: "claude-opus-5", effortName: "max" },
+  "opus-5-med": { model: 6, effort: 1, modelId: "claude-opus-5-5", effortName: "medium" },
+  "opus-5-high": { model: 6, effort: 2, modelId: "claude-opus-5-5", effortName: "high" },
+  "opus-5-xhigh": { model: 6, effort: 3, modelId: "claude-opus-5-5", effortName: "xhigh" },
+  "opus-5-max": { model: 6, effort: 4, modelId: "claude-opus-5-5", effortName: "max" },
   "fable-5-med": { model: 7, effort: 1, modelId: "claude-fable-5", effortName: "medium" },
   "fable-5-high": { model: 7, effort: 2, modelId: "claude-fable-5", effortName: "high" },
   "fable-5-xhigh": { model: 7, effort: 3, modelId: "claude-fable-5", effortName: "xhigh" },
@@ -5234,15 +5256,15 @@ function diyValidate(cfg) {
   if (cfg.autonomy === "hands-off" && (cfg.ship_mode === "commit" || cfg.ship_mode === "pr")) {
     warnings.push(`hands-off + ship_mode ${cfg.ship_mode}: git actions will run fully unattended`);
   }
-  // The pinned reviewer/verifier moved to claude-opus-5 (model rank 6) in
+  // The pinned reviewer/verifier moved to claude-opus-5-5 (model rank 6) in
   // v0.34.0 — anything below that tier silently runs them at the session model.
   if (tier && tier.model < 6 && (cfg.review !== "off" || cfg.verify !== "off")) {
     // Careful wording (v0.34.7): `session_tier` is a DECLARATION, not the real
     // session model — a hook cannot read the model, only the effort. On a
-    // session that actually outranks the declared tier the pinned Opus 5 roles
+    // session that actually outranks the declared tier the pinned Opus 5.5 roles
     // run at FULL pin (observed), so asserting they "will" downgrade is a false
     // alarm; what is certain is only that a LOWER real session caps them.
-    warnings.push(`session_tier ${cfg.session_tier} is below the pinned Opus 5 reviewer/verifier: if the REAL session is also below Opus 5 they run capped at it (the tier-honesty rule reports the actual model). A session above this tier runs them at full pin — but your executor table stays clipped to ${cfg.session_tier}; recompile to use the full ladder`);
+    warnings.push(`session_tier ${cfg.session_tier} is below the pinned Opus 5.5 reviewer/verifier: if the REAL session is also below Opus 5.5 they run capped at it (the tier-honesty rule reports the actual model). A session above this tier runs them at full pin — but your executor table stays clipped to ${cfg.session_tier}; recompile to use the full ladder`);
   }
   return { errors, warnings };
 }
@@ -7626,7 +7648,7 @@ const WIKI_TIER_LADDER = [
   { id: "small-delta", tier: "light", why: "small delta, no new surface" },
 ];
 
-// The POSITION each tier resolves to (v0.55.0). Two slots and ONE Opus 5 agent
+// The POSITION each tier resolves to (v0.55.0). Two slots and ONE Opus 5.5 agent
 // is not a contradiction: a slot names the POSITION, not the model, and the
 // ladder already collapses both tiers onto that agent while `opus5_only` is on.
 const WIKI_TIER_SLOT = { deep: "wiki-scanner-deep", light: "wiki-scanner-light" };
@@ -8742,6 +8764,14 @@ function gainRowFor(sub, Q, model, operands, r, ms, forSlice) {
     paid.source = K.tok(r.source.text);
     paid.card = Math.max(0, paid.card - paid.source);
   }
+  // R1 (v1.9.1): the callers' blocks are source the same way — exactly the
+  // lines a range read of each caller would have returned.
+  const callerBlocks = Array.isArray(r.callers_source) ? r.callers_source : [];
+  if (callerBlocks.length) {
+    const t = callerBlocks.reduce((a, b) => a + K.tok(b.text || ""), 0);
+    paid.source += t;
+    paid.card = Math.max(0, paid.card - t);
+  }
   const base = { cmd: sub, targets: operands.slice(0, 5), gen, ms, paid };
 
   if (sub === "impact" || (sub === "ctx" && forSlice)) {
@@ -8773,6 +8803,7 @@ function gainRowFor(sub, Q, model, operands, r, ms, forSlice) {
     const hits = Q.nameHits(model, r.target.qname);
     const c = K.ctxSymbol(model, r.target, hits, (r.callers || []).map((x) => x.file));
     if (r.source) c.avoided = addAvoided(c.avoided, K.sourceAvoided(model, r.source.file, r.source.from, r.source.to));
+    for (const b of callerBlocks) c.avoided = addAvoided(c.avoided, K.sourceAvoided(model, b.file, b.from, b.to));
     return { ...base, ...c };
   }
   return { ...base, avoided: { low: 0, high: 0, calls_low: 0, calls_high: 0 }, basis: { files: [], grep_hits: 0 } };
@@ -8815,23 +8846,79 @@ function graphGainAppend(claudeDir, row) {
 // A thin router. The engine is `bin/graph.js`; this function resolves the
 // config (the engine never reads it) and owns the human lines. The exit code is
 // the same on the human path and the `--json` path (S7).
+// ── B1 (v1.9.1): the BRIEF answer ─────────────────────────────────────────
+// The orchestrator reads a graph answer on the main-session surface, where it
+// is re-sent on every later turn — and it prints the `card`, never the rows.
+// Measured on this tree the rows are 4 to 8 times the card. `--brief` keeps
+// every fact a lane prints and drops every row array it does not.
+//
+// Two rules, in this order:
+//   1. an array NAMED here is dropped (`calls[]` and `names[]` are strings, and
+//      they are still rows);
+//   2. any other array OF OBJECTS is dropped, unless it IS the answer
+//      (`nearest[]` and `candidates[]` on an exit-4; `cards[]` on a multi-target
+//      `ctx`, whose rows are three scalars each).
+// Every drop leaves its COUNT behind, at the level it was dropped from, except
+// where the answer already carries that count under another name.
+const GRAPH_BRIEF_DROP = new Set(["callers", "calls", "effects", "symbols", "imports", "importers", "blocks", "names", "rows", "chain", "waiting"]);
+const GRAPH_BRIEF_KEEP = new Set(["nearest", "candidates", "cards"]);
+// V2/V3 answers that ARE the answer: `complexity` and `facts` are ten rows at
+// most, and they are the whole reason the call was made. They pass through
+// whole, never walked.
+const GRAPH_BRIEF_WHOLE = new Set(["complexity", "facts"]);
+const GRAPH_BRIEF_NO_COUNT = { impact: ["callers"], map: ["files"], changes: ["symbols"] };
+
+function briefOf(sub, obj) {
+  const noCount = new Set(GRAPH_BRIEF_NO_COUNT[sub] || []);
+  const walk = (o) => {
+    const out = {};
+    const counts = {};
+    for (const [k, v] of Object.entries(o)) {
+      if (Array.isArray(v)) {
+        const rows = GRAPH_BRIEF_DROP.has(k) || (!GRAPH_BRIEF_KEEP.has(k) && v.length > 0 && v[0] !== null && typeof v[0] === "object");
+        if (!rows) out[k] = v;
+        else if (!noCount.has(k)) counts[k] = v.length;
+        continue;
+      }
+      if (k === "source" && v && typeof v === "object") {
+        const from = Number(v.from);
+        const to = Number(v.to);
+        counts.source_lines = Number.isFinite(from) && Number.isFinite(to) ? Math.max(0, to - from + 1) : 0;
+        continue;
+      }
+      out[k] = v && typeof v === "object" && !GRAPH_BRIEF_WHOLE.has(k) ? walk(v) : v;
+    }
+    if (Object.keys(counts).length) out.counts = { ...(out.counts || {}), ...counts };
+    return out;
+  };
+  return walk(obj);
+}
+
 function graphCmd() {
   const usage =
-    "Usage: orc graph status [--heal] | update [--notes-pending --files <a,b>] | gc\n" +
-    "       orc graph ctx <symbol|file[:line]>… [--source [N]] | ctx --for-slice <file…> | impact <file…> | path <from> <to>\n" +
-    "       orc graph coverage <file…> | changes [--base <ref>] | cochange <file>\n" +
+    "Usage: orc graph status [--heal] | update [--notes-pending --files <a,b>] | gc | audit [--top=N]\n" +
+    "       orc graph ctx <symbol|file[:line]>… [--source [N]] [--callers-source] | ctx --for-slice <file…> | impact <file…> | path <from> <to>\n" +
+    "       orc graph coverage <file…> | changes [--base <ref>] [--files=<a,b>] | cochange <file>\n" +
     "       orc graph map [--focus <file|name,…>] [--budget N]\n" +
     "       orc graph gain [--run <name>] [--since 7d] [--history [--limit 40]] [--measured] [--reset --yes]\n" +
     "       orc graph notes pending --files <a,b> [--cap 40] [--min 5] [--with-source] | notes apply <file|->\n" +
-    "       [--if-enabled] [--depth N] [--budget TOKENS] [--format prose|tree] [--offset N] [--json] [--dir <path>]\n" +
+    "       [--if-enabled] [--depth N] [--budget TOKENS] [--format prose|tree] [--offset N] [--json] [--brief] [--dir <path>]\n" +
+    "  --brief  with --json: the card, the line, the trace, every scalar and every count — and none of the row\n" +
+    "           arrays a lane never prints. Compact output. Without --brief, --json is unchanged.\n" +
     "  status   exit 0 FRESH · 1 NONE · 2 DRIFTED · 3 OFF  (--heal builds or updates a NONE/DRIFTED graph first)\n" +
     "  update   exit 0 done · 1 unavailable or locked · 3 off (with --if-enabled)\n" +
     "           --notes-pending adds the `notes pending` answer to the SAME call (one process, one lock)\n" +
     "  gc       exit 0 done · 1 no index or locked\n" +
+    "  audit    exit 0 answered · 1 no index · 3 off (with --if-enabled)   WHY the map is thin: density by language,\n" +
+    "           the files the parser read as EMPTY, longest first, and what their first declaration looks like\n" +
     "  ctx      exit 0 found · 1 no graph · 3 off (with --if-enabled) · 4 not found or ambiguous\n" +
     "           --source [N] appends the target's own lines (default 80, cap 200), charged to the same budget\n" +
+    "           --callers-source appends 6 lines around each CONFIDENT call site (≤ 5, never the target's own\n" +
+    "           file), charged to the same budget after everything else; symbol cards only\n" +
     "           --for-slice prints only the OUTSIDE view of each declared file — no symbol table\n" +
     "  impact   exit 0 found · 1 no graph · 3 off (with --if-enabled) · 4 none of the files is in the graph\n" +
+    "           --complexity [--risk=<class>[@<file:line>],…] adds the ONE line that says whether this change\n" +
+    "           fits one executor, with the numbers behind it — impact and co-change in ONE call\n" +
     "  path     exit 0 found · 1 no graph · 3 off (with --if-enabled) · 4 no confident path, or an unknown end\n" +
     "  map      exit 0 always when a graph exists · 1 no graph · 3 off (with --if-enabled)   the files by RANK, most\n" +
     "           connected first, each with its top symbols; --focus re-ranks the repository around those files or names\n" +
@@ -8852,6 +8939,16 @@ function graphCmd() {
   const root = repoRootOf(claudeDir);
   const sub = positionals()[1];
   const asJson = wantsJson();
+  // B1: a BARE switch, read from `args`. `flag()` reads the next word as a
+  // value, so `--brief` there would swallow the first operand.
+  let briefOn = args.includes("--brief");
+  // Every flag this release adds that carries a VALUE is written
+  // `--name=value`, for the same reason: one word, so a CLI that does not
+  // know the flag skips it whole instead of losing an operand behind it.
+  const eqFlag = (name) => {
+    const hit = args.find((a) => a.startsWith(name + "="));
+    return hit === undefined ? undefined : hit.slice(name.length + 1);
+  };
   const ovr = readOverride(claudeDir).map;
   const cfg = (k) => (Object.prototype.hasOwnProperty.call(ovr, k) ? ovr[k] : metaFor(k).def);
   const enabled = String(cfg("code_graph")) === "on";
@@ -8930,7 +9027,12 @@ function graphCmd() {
       return null;
     }
   };
-  const finish = (r, line, trace, jsonLine = true) => {
+  // B3 (v1.9.1): the ledger append moved INSIDE `finish`, because the one
+  // number the meter called EXACT — the card — was 4 to 8 times too small. The
+  // ENVELOPE is everything the answer carries beyond its card, and it can only
+  // be counted once the text that will be written exists. The append is still
+  // fail-quiet and still happens after the answer is built, never before it.
+  const finish = (r, line, trace, jsonLine = true, gainRow = null) => {
     const { exit, ...obj } = r;
     if (obj.generation === undefined) {
       const m = metaNow();
@@ -8939,7 +9041,19 @@ function graphCmd() {
         obj.gen_id = m.gen_id || null;
       }
     }
-    if (asJson) emitJson({ ...obj, ...(line && jsonLine ? { line } : {}), ...(trace ? { trace } : {}) }, exit);
+    const full = { ...obj, ...(line && jsonLine ? { line } : {}), ...(trace ? { trace } : {}) };
+    const payload = asJson && briefOn ? briefOf(sub, full) : full;
+    if (gainRow) {
+      try {
+        const K = require("./graph-gain.js");
+        const text = asJson ? (briefOn ? JSON.stringify(payload) : JSON.stringify(payload, null, 2)) : String(line || "");
+        if (gainRow.paid) gainRow.paid.envelope = Math.max(0, K.tok(text) - K.tok(obj.card || line || ""));
+        graphGainAppend(claudeDir, gainRow);
+      } catch (_) {
+        // K6: the meter is never allowed to fail the read it is measuring.
+      }
+    }
+    if (asJson) emitJson(payload, exit, briefOn);
     if (line) console.log(line);
     process.exit(exit);
   };
@@ -8962,6 +9076,32 @@ function graphCmd() {
     r.auto_update = autoUpdate;
     r.notes = String(cfg("code_graph_notes"));
     if (healed) r.healed = healed;
+    // A1 (v1.9.1): the density, and the BACKFILL for a store built before
+    // this release. A FRESH store never updates, so waiting for the next real
+    // update would leave a long-lived graph without the line forever. It runs
+    // under `--heal` only, once, and it is bounded by the same cap every heal
+    // obeys — the index did not change, so `generation` and `gen_id` do not.
+    const SIGD = require("./graph-signals.js");
+    let meta = metaNow();
+    if (flag("--heal") === true && enabled && meta && !meta.density && r.exists) {
+      const capMs = Number(cfg("code_graph_heal_ms")) || 0;
+      if (!capMs || !meta.update_ms || meta.update_ms <= capMs) {
+        try {
+          const p = G.graphPaths(claudeDir);
+          const idx = JSON.parse(fs.readFileSync(p.index, "utf8"));
+          const fl = JSON.parse(fs.readFileSync(p.files, "utf8"));
+          meta.density = G.densityOf(idx, fl);
+          G.atomicWrite(p.meta, JSON.stringify(meta, null, 2) + "\n");
+          r.density_backfilled = true;
+        } catch (_) {
+          meta = metaNow();
+        }
+      }
+    }
+    if (meta && meta.density) {
+      r.density = meta.density;
+      r.thin = SIGD.isThin(meta.density, r.files);
+    }
     const b = r.behind || {};
     const lines = {
       off: "graph: off",
@@ -8975,6 +9115,12 @@ function graphCmd() {
     };
     let line = lines[r.state];
     let consult = r.state;
+    // The density rides on the line a lane already prints, never on a line of
+    // its own: a second line is a second thing to skip.
+    const densityText = r.density
+      ? ` · ${r.density.symbols_per_file} symbols/file` +
+        (r.thin ? ` · THIN (${Math.round(r.density.zero_share * 100)} % of files have no symbol) — run: orc graph audit` : "")
+      : "";
     if (healed && healed.parsed !== undefined) {
       const moved = healed.added + healed.changed + healed.deleted;
       const size = `${plural(r.files, "file")} · ${plural(r.symbols, "symbol")}`;
@@ -8986,7 +9132,15 @@ function graphCmd() {
             ? `graph: index upgraded to the new engine → updated (${healed.ms} ms) — ${size}`
             : `graph: ${plural(moved, "file")} changed outside ORC → updated (${healed.ms} ms) — ${size}`;
     } else if (healed) line += ` · could not heal: ${healed.reason}`;
-    return finish(r, line, consult === "off" ? "GRAPH-CONSULT off" : `GRAPH-CONSULT ${consult} :: files=${r.files} symbols=${r.symbols} gen=${r.generation}`);
+    if (densityText && ["fresh", "drifted", "built", "updated"].includes(consult)) line += densityText;
+    return finish(
+      r,
+      line,
+      consult === "off"
+        ? "GRAPH-CONSULT off"
+        : `GRAPH-CONSULT ${consult} :: files=${r.files} symbols=${r.symbols} gen=${r.generation}` +
+          (r.density ? ` density=${r.density.symbols_per_file}${r.thin ? " thin=1" : ""}` : "")
+    );
   }
 
   if (sub === "update") {
@@ -9150,8 +9304,11 @@ function graphCmd() {
     const SIG = require("./graph-signals.js");
     if (sub === "changes") {
       const base = typeof flag("--base") === "string" ? flag("--base") : null;
+      // V1: the lane hands the files it just wrote, and the rows for every
+      // other file never leave this process.
+      const onlyFiles = String(eqFlag("--files") || "").split(",").map((x) => x.trim()).filter(Boolean);
       const healed = healOnRead(SIG.changedPaths(root, base));
-      const r = SIG.graphChanges(claudeDir, root, { base });
+      const r = SIG.graphChanges(claudeDir, root, { base, files: onlyFiles });
       if (healed) r.healed = healed;
       if (r.exit !== 0) return finish(r, r.reason === "no-index" ? "graph: none — no index yet (run: orc graph update)" : `graph: unavailable — ${r.reason}`);
       const c = r.counts;
@@ -9163,6 +9320,7 @@ function graphCmd() {
         : `graph changes vs ${r.base} — no indexed symbol was touched${r.not_in_graph.length ? ` (changed, but not in the graph: ${r.not_in_graph.join(", ")})` : ""}`;
       // K1/K2: `changes` avoids one Grep per touched symbol. The diff itself
       // is NOT counted — the reviewer reads it either way.
+      let gainRow = null;
       if (r.symbols.length) {
         try {
           const K = require("./graph-gain.js");
@@ -9171,11 +9329,11 @@ function graphCmd() {
           if (m) {
             const hits = r.symbols.reduce((a, x) => a + Q.nameHits(m, x.qname), 0);
             const callerFiles = r.symbols.flatMap((x) => (x.callers || []).map((cc) => cc.file || cc));
-            graphGainAppend(claudeDir, { cmd: "changes", targets: [r.base], gen: r.generation, ms: r.ms || 0, paid: { card: K.tok(line), source: 0, hints: 0 }, ...K.changesCost(m, r.symbols.length, hits, callerFiles) });
+            gainRow = { cmd: "changes", targets: [r.base], gen: r.generation, ms: r.ms || 0, paid: { card: K.tok(line), source: 0, hints: 0 }, ...K.changesCost(m, r.symbols.length, hits, callerFiles) };
           }
         } catch (_) {}
       }
-      return finish(r, line, `GRAPH-CHANGES ${r.symbols.length ? "found" : "none"} :: symbols=${r.symbols.length} high=${c.high} medium=${c.medium} low=${c.low} gen=${r.generation}`);
+      return finish(r, line, `GRAPH-CHANGES ${r.symbols.length ? "found" : "none"} :: symbols=${r.symbols.length} high=${c.high} medium=${c.medium} low=${c.low} gen=${r.generation}`, true, gainRow);
     }
     const file = plain[2];
     if (!file) return finish({ ok: false, reason: "missing-operand", usage, exit: 1 }, usage);
@@ -9190,6 +9348,18 @@ function graphCmd() {
     return finish(r, line, `GRAPH-COCHANGE ${r.rows.length ? "found" : "none"} :: rows=${r.rows.length} commits=${r.commits}`);
   }
 
+  // A2 (v1.9.1) — `orc graph audit`. A THIN line says the map is thin; this
+  // says WHY, and it is a USER command: no lane runs it (`LANE_CALLS` pins
+  // `lanes: []`). It reads the index, the file table and the first lines of
+  // each empty file. It never parses, never writes, and costs no model tokens.
+  if (sub === "audit") {
+    if (!enabled && flag("--if-enabled") === true) return finish({ ok: true, enabled: false, state: "off", exit: 3 }, "graph: off");
+    const A = require("./graph-audit.js");
+    const r = A.audit(claudeDir, root, { top: eqFlag("--top") });
+    if (r.exit === 1) return finish(r, "graph: none — no index yet (run: orc graph update)");
+    return finish(r, r.line, r.trace, false);
+  }
+
   if (sub === "gc") {
     const r = G.graphGc(claudeDir);
     const line =
@@ -9289,16 +9459,21 @@ function graphCmd() {
     if (flag("--history") !== undefined) {
       const r = K.history(claudeDir, { limit: flag("--limit") });
       if (!r.ok) return finish(r, "graph gain: no calls recorded yet");
+      // `--history` IS its rows. Dropping them would leave an answer with
+      // nothing in it, so the switch is refused here — out loud.
+      const briefAsked = briefOn;
+      briefOn = false;
       const line =
         `graph gain — last ${r.rows.length} of ${r.total} call(s), newest first\n` +
         r.rows
           .map(
             (x) =>
               `  ${x.at}  ${String(x.cmd).padEnd(10)} ${String((x.targets || []).join(",")).slice(0, 40).padEnd(40)}` +
-              ` paid ${String(x.paid.card + x.paid.source + x.paid.hints).padStart(6)}  avoided ~${rng(x.avoided.low, x.avoided.high)}`
+              ` paid ${String(x.paid.card + x.paid.source + x.paid.hints + (x.paid.envelope || 0)).padStart(6)}  avoided ~${rng(x.avoided.low, x.avoided.high)}`
           )
           .join("\n") +
-        "\n  `avoided` is an ESTIMATE of retrieval the ladder would have paid for — never a bill";
+        "\n  `avoided` is an ESTIMATE of retrieval the ladder would have paid for — never a bill" +
+        (briefAsked ? " (--brief ignored on --history)" : "");
       return finish(r, line);
     }
 
@@ -9318,11 +9493,21 @@ function graphCmd() {
     const scope = r.run ? `run ${r.run}` : r.since ? `last ${r.since}` : "this project";
     const line =
       U.header(`graph gain — ${scope} · ${plural(r.calls_recorded, "call")} · ${plural(r.runs, "run")}${g ? ` · gen ${g[0]}${g[1] !== g[0] ? `–${g[1]}` : ""}` : ""}`) +
-      `\n  paid       ${kTok(r.paid.total).padStart(10)} tokens   (cards ${kTok(r.paid.card)} · source ${kTok(r.paid.source)} · hints ${kTok(r.paid.hints)})` +
+      `\n  paid       ${kTok(r.paid.total).padStart(10)} tokens   (cards ${kTok(r.paid.card)} · source ${kTok(r.paid.source)} · hints ${kTok(r.paid.hints)} · envelope ${kTok(r.paid.envelope)})` +
       `\n  avoided    ~${rng(r.avoided.low, r.avoided.high)} tokens   (estimate — the read ladder done well … done badly)` +
       `\n  net        ~${rng(r.net.low, r.net.high)} tokens · ~${r.calls.low} – ${r.calls.high} tool calls` +
       (byCmd ? `\n  by command ${byCmd}` : "") +
       `\n  hints      ${r.hints.injected} injected · ${r.hints.read_notes} read notes · ${r.hints.updates} hook updates` +
+      // R2 (v1.9.1): the reads the `on,read` hint would have named ranges for.
+      // A COUNT, printed only when there is one — it is the data DE-8 waits
+      // for, and nothing changes a setting because of it.
+      (r.hints.wide_unhinted
+        ? `\n  wide reads   ${r.hints.wide_unhinted} whole-file reads of files with ${require("./graph-resolve.js").WIDE_MIN}+ symbols were not hinted — \`code_graph_hooks: on,read\` names their ranges`
+        : "") +
+      // M2: a read nobody made is a fact about this project, printed with no
+      // advice after it. The graph is not wrong for holding an answer nobody
+      // asked for, and the meter does not tell anyone to start asking.
+      (r.never_called && r.never_called.length ? `\n  never called  ${r.never_called.join(" · ")}` : "") +
       (r.unsized ? `
   ${plural(r.unsized, "row")} answered from SHARDS, where the size of the search replaced cannot be known — their estimate leaves it out` : "") +
       "\n  an estimate of avoided RETRIEVAL, never a bill — searches are a fraction of a percent of a session" +
@@ -9365,8 +9550,8 @@ function graphCmd() {
     // gets it too. Appending it here would have shown it only to a human.
     const line = r.card;
     const trace = `GRAPH-MAP ${focus.length ? "focused" : "repo"} :: files=${r.shown}/${r.total_files}${focus.length ? ` focus=${r.focus.join(",")}` : ""} gen=${r.generation}`;
-    if (r.exit === 0) graphGainAppend(claudeDir, gainRowFor("map", Q, model, focus, r, Date.now() - t0, false));
-    return finish(r, line, trace, false);
+    const gainRow = r.exit === 0 ? gainRowFor("map", Q, model, focus, r, Date.now() - t0, false) : null;
+    return finish(r, line, trace, false, gainRow);
   }
 
   if (sub === "ctx" || sub === "impact" || sub === "path") {
@@ -9391,6 +9576,8 @@ function graphCmd() {
     // D2 is a SWITCH, so it is read from `args` — `flag()` would read the first
     // declared file as its value.
     const forSlice = args.includes("--for-slice");
+    // R1 (v1.9.1): a SWITCH too — six lines around each confident call site.
+    const callersSource = args.includes("--callers-source");
     // `path` names two symbols, not files, so it has nothing cheap to check —
     // it reads whatever generation is on disk, like every other command did
     // before EW3.
@@ -9411,7 +9598,7 @@ function graphCmd() {
     let fast = null;
     if (sub === "ctx" && !forSlice && operands.length === 1) {
       try {
-        fast = require("./graph-shard.js").fastModel(claudeDir, root, operands[0]);
+        fast = require("./graph-shard.js").fastModel(claudeDir, root, operands[0], { callersSource });
       } catch (_) {
         fast = null;
       }
@@ -9439,6 +9626,7 @@ function graphCmd() {
       format: String(flag("--format") || "prose") === "tree" ? "tree" : "prose",
       offset: flag("--offset"),
       source: sourceLines,
+      callersSource,
       wikiDocsFor,
     };
     let r;
@@ -9498,11 +9686,22 @@ function graphCmd() {
     } else if (sub === "impact") {
       if (!operands.length) return finish({ ok: false, reason: "missing-operand", usage, exit: 1 }, usage);
       r = Q.impact(model, operands, opts);
+      // V2/V3 (v1.9.1): the complexity read, and the `graph_facts` the
+      // planner pastes — computed here, in ONE process, from the model this
+      // call already loaded. Mini used to run `impact` plus one `cochange`
+      // per declared file and count the rows in its own context.
+      if (args.includes("--complexity") && r.exit === 0) {
+        const SIG = require("./graph-signals.js");
+        const cx = SIG.complexityOf(claudeDir, root, model, r.files, { risk: eqFlag("--risk"), tests: r.tests });
+        r.complexity = { verdict: cx.verdict, why: cx.why, numbers: cx.numbers, line: cx.line, thresholds: cx.thresholds };
+        r.facts = cx.facts;
+      }
     } else {
       if (operands.length < 2) return finish({ ok: false, reason: "missing-operand", usage, exit: 1 }, usage);
       r = Q.pathBetween(model, operands[0], operands[1]);
     }
     let line = r.card;
+    if (line && r.complexity) line = `${line}\n  ${r.complexity.line}`;
     if (!line && r.state === "not-found") {
       const near = (r.nearest || []).map((s) => `${s.qname} (${s.file}:${s.lines[0]})`);
       line = r.missing
@@ -9526,21 +9725,26 @@ ${line || ""}`;
 ` +
         (line || "");
     }
+    const cx = r.complexity || null;
     const trace =
-      sub === "ctx" && r.exit === 0
+      cx
+        ? `GRAPH-COMPLEXITY ${cx.verdict} :: files=${cx.numbers.files} callers=${cx.numbers.callers_outside} caller_files=${cx.numbers.caller_files_outside}` +
+          ` maybe=${cx.numbers.maybe} tests=${cx.numbers.tests_reaching} risk=${cx.numbers.risk.length} cochange=${cx.numbers.cochange_missing.length} gen=${r.generation || (r.facts && r.facts.generation) || 0}`
+        : sub === "ctx" && r.exit === 0
         ? `GRAPH-CONSULT ${r.view === "for-slice" ? "slice" : "card"} :: targets=${(r.view === "for-slice" ? r.files : r.cards ? r.cards.filter((c) => c.state === "found").map((c) => c.query) : [operands[0]]).join(",")}` +
           (healed && healed.state && healed.state !== "skipped" ? ` healed=${healed.trigger}` : "")
         : null;
+    let gainRow = null;
     if (r.exit === 0) {
       try {
-        graphGainAppend(claudeDir, gainRowFor(sub, Q, model, operands, r, Date.now() - t0, forSlice));
+        gainRow = gainRowFor(sub, Q, model, operands, r, Date.now() - t0, forSlice);
       } catch (_) {
         // K6: the meter is never allowed to fail the read it is measuring. The
         // append was already fail-quiet; the PRICING was not, because it was an
         // argument to it.
       }
     }
-    return finish(r, line, trace, !r.card);
+    return finish(r, line, trace, !r.card, gainRow);
   }
 
   if (asJson) emitJson({ ok: false, reason: "unknown-subcommand", usage }, 1);
@@ -11034,7 +11238,7 @@ function resume() {
 // assumed the opposite, so an interrupted turn produced a SECOND agent on the
 // same task while the first was still working. A graded run put THREE
 // `orc-executor-opus-5-low` agents on one task for 50m19s + 115m22s + 100m53s
-// — 266 minutes of Opus 5 for one authorised dispatch, all editing the same
+// — 266 minutes of Opus 5.5 for one authorised dispatch, all editing the same
 // files. The hook already recorded every one of them; nothing ever READ it.
 //
 // The pending sidecar (`<trace>.pending.json`, written by orc-trace.js on every
@@ -21038,7 +21242,7 @@ function docForecastCmd(claudeDir, slugArg) {
     for (const part of roleParts) {
       const vec = vecFor(part.agent, part.n, pct);
       if (!part.target) {
-        const m = priceVector(claudeDir, vec, "claude-opus-5");
+        const m = priceVector(claudeDir, vec, "claude-opus-5-5");
         if (m.usd === null) priceable = false;
         else total += m.usd;
         continue;
@@ -21062,8 +21266,8 @@ function docForecastCmd(claudeDir, slugArg) {
   };
   const usd50 = usdAt("p50");
   const usd90 = usdAt("p90");
-  const money50 = priceVector(claudeDir, p50, "claude-opus-5");
-  const money90 = priceVector(claudeDir, p90, "claude-opus-5");
+  const money50 = priceVector(claudeDir, p50, "claude-opus-5-5");
+  const money90 = priceVector(claudeDir, p90, "claude-opus-5-5");
   const extraView = {
     resolved: docExtra.resolved,
     why: docExtra.why,
@@ -21282,7 +21486,7 @@ function docCostCmd(claudeDir, slugArg) {
     runRows.push({ trace: r.name, lane: r.lane, date: r.date, tokens: runVec, weighted: weightedTokens(runVec) });
   }
 
-  const money = priceVector(claudeDir, total, "claude-opus-5");
+  const money = priceVector(claudeDir, total, "claude-opus-5-5");
   const sections = Object.values(bySection).map((s) => ({
     id: s.id,
     heading: s.heading,
@@ -21291,7 +21495,7 @@ function docCostCmd(claudeDir, slugArg) {
     joined: s.joined > 0,
     tokens: s.joined ? s.vec : null,
     weighted: s.joined ? weightedTokens(s.vec) : null,
-    usd: s.joined ? priceVector(claudeDir, s.vec, "claude-opus-5").usd : null,
+    usd: s.joined ? priceVector(claudeDir, s.vec, "claude-opus-5-5").usd : null,
   }));
   const unjoined = sections.filter((s) => !s.joined).length;
   const code = joined === 0 ? 3 : unjoined ? 1 : 0;
@@ -25625,8 +25829,8 @@ function extraResolveFor(claudeDir, score, opts) {
 //
 // `claude` is an ARRAY because `quick-executor` has two: a menu is what that
 // lane is. `claude_opus5` is the variant `opus5_only` would have used, derived
-// from the shipped ladders — NULL where the agent is already Opus 5, so two
-// wiki slots collapsing onto one Opus 5 scanner is not a contradiction: the
+// from the shipped ladders — NULL where the agent is already Opus 5.5, so two
+// wiki slots collapsing onto one Opus 5.5 scanner is not a contradiction: the
 // slot names the POSITION, not the model. NO AGENT IS ADDED by this release
 // and no pair is needed (the floor stays 51).
 const EXTRA_SLOTS = [
@@ -43401,7 +43605,7 @@ function slEngine() {
 // panel really hands over — including `tokenCount`, which is the number v1.2.0
 // concluded could not be measured.
 const SL_TASK_FIXTURES = {
-  healthy: { id: "t1", name: "orc-executor-opus-5-low", type: "orc-executor-opus-5-low", description: "wire the retry ladder", status: "running", model: "claude-opus-5", effort: "low", tokenCount: 84000, contextWindowSize: 200000, startTime: 1767225600000 - 17 * 60000 },
+  healthy: { id: "t1", name: "orc-executor-opus-5-low", type: "orc-executor-opus-5-low", description: "wire the retry ladder", status: "running", model: "claude-opus-5-5", effort: "low", tokenCount: 84000, contextWindowSize: 200000, startTime: 1767225600000 - 17 * 60000 },
   degraded: { id: "t2", name: "orc-executor-sonnet-4-6-med", type: "orc-executor-sonnet-4-6-med", description: "rename two files", status: "failed", model: "claude-sonnet-4-6", effort: "medium", tokenCount: 191000, contextWindowSize: 200000, startTime: 1767225600000 - 96 * 60000 },
   empty: { id: "t3" },
 };
@@ -43410,7 +43614,7 @@ const SL_FIXTURES = {
   healthy: {
     label: "a healthy session mid-run",
     payload: {
-      model: { id: "claude-opus-5", display_name: "Opus 5" },
+      model: { id: "claude-opus-5-5", display_name: "Opus 5.5" },
       effort: { level: "high" },
       version: "2.1.80",
       context_window: { used_percentage: 38, remaining_percentage: 62, context_window_size: 200000 },