pi-smart-router 0.7.0 → 0.9.0

package/.pi/extensions/smart-router/command-formatters.ts

@@ -2,7 +2,11 @@ import {
   DEFAULT_HISTORY_LIMIT,
   MAX_HISTORY_LIMIT,
 } from '../../../src/infrastructure/telemetry/telemetry-limits.js';
-import type { RoutingDecision, RoutingTelemetry } from '../../../src/domain/types/index.js';
+import type {
+  ModelProfile,
+  RoutingDecision,
+  RoutingTelemetry,
+} from '../../../src/domain/types/index.js';
 import { SMART_ROUTER_USAGE } from './commands.js';
 import {
   DEFAULT_TELEMETRY_CONTRIB_EXPORT_LIMIT,
@@ -15,6 +19,78 @@ import {
 import { formatPricingStalenessLine } from './pricing-lifecycle.js';
 import type { FleetMode, SmartRouterCommand, SmartRouterRuntime } from './types.js';
 
+/** Opaque / virtual auto ids that hide the concrete delegated fleet model (SP-178). */
+function isBareOrSmartRouterAuto(modelId: string): boolean {
+  return modelId === 'auto' || modelId === 'smart-router/auto';
+}
+
+/**
+ * Resolve the operator-facing model id for history/status (SP-178 / #99).
+ * Prefer a concrete delegated/primary id over virtual `auto`.
+ */
+export function resolveHistoryModelId(
+  entry: Pick<
+    RoutingTelemetry,
+    | 'selected_model_id'
+    | 'planning_delegate_primary_model_id'
+    | 'planning_delegate_model_id'
+  >,
+  fleet?: readonly ModelProfile[],
+): string {
+  let modelId = entry.selected_model_id;
+
+  if (isBareOrSmartRouterAuto(modelId)) {
+    const primary = entry.planning_delegate_primary_model_id;
+    if (primary && !isBareOrSmartRouterAuto(primary)) {
+      modelId = primary;
+    } else if (
+      entry.planning_delegate_model_id &&
+      !isBareOrSmartRouterAuto(entry.planning_delegate_model_id)
+    ) {
+      modelId = entry.planning_delegate_model_id;
+    }
+  }
+
+  return qualifyModelIdForDisplay(modelId, fleet);
+}
+
+/**
+ * Qualify bare `auto` with provider when fleet is available so history never
+ * looks like the smart-router virtual model.
+ */
+export function qualifyModelIdForDisplay(
+  modelId: string,
+  fleet?: readonly ModelProfile[],
+): string {
+  if (modelId !== 'auto') {
+    return modelId;
+  }
+
+  const profile = fleet?.find((m) => m.id === 'auto');
+  if (profile) {
+    return `${profile.provider}/${profile.id}`;
+  }
+
+  // Cursor opaque auto is the common bare-`auto` fleet id; never leave it unqualified.
+  return 'cursor/auto';
+}
+
+export function resolveStatusModelId(
+  decision: RoutingDecision,
+  fleet?: readonly ModelProfile[],
+): string {
+  const primary = decision.features?.planning_delegate?.primary_model_id ?? null;
+  const delegate = decision.features?.planning_delegate?.delegate_model_id ?? null;
+  return resolveHistoryModelId(
+    {
+      selected_model_id: decision.selected_model_id,
+      planning_delegate_primary_model_id: primary,
+      planning_delegate_model_id: delegate,
+    },
+    fleet,
+  );
+}
+
 export function parseHistoryLimit(raw: string | undefined): number {
   if (raw === undefined) {
     return DEFAULT_HISTORY_LIMIT;
@@ -142,8 +218,9 @@ export function formatStatusMessage(
     return lines.join('\n');
   }
 
+  const displayModelId = resolveStatusModelId(decision, runtime.streamDeps.fleet);
   lines.push(
-    `Model: ${decision.selected_model_id}`,
+    `Model: ${displayModelId}`,
     `Stage: ${decision.stage}`,
     `Reason: ${decision.reason_code}`,
     `Latency: ${decision.routing_latency_ms}ms`,
@@ -151,16 +228,20 @@ export function formatStatusMessage(
   return lines.join('\n');
 }
 
-export function formatHistoryMessage(entries: readonly RoutingTelemetry[]): string {
+export function formatHistoryMessage(
+  entries: readonly RoutingTelemetry[],
+  options?: { fleet?: readonly ModelProfile[] },
+): string {
   if (entries.length === 0) {
     return 'No routing history yet.';
   }
 
+  const fleet = options?.fleet;
   return entries
-    .map(
-      (entry) =>
-        `${entry.timestamp} | ${entry.selected_model_id} | ${entry.stage} | ${entry.turn_type} | ${entry.routing_latency_ms}ms`,
-    )
+    .map((entry) => {
+      const modelId = resolveHistoryModelId(entry, fleet);
+      return `${entry.timestamp} | ${modelId} | ${entry.stage} | ${entry.turn_type} | ${entry.routing_latency_ms}ms`;
+    })
     .join('\n');
 }
 

package/.pi/extensions/smart-router/commands.ts

@@ -160,7 +160,10 @@ export function registerSmartRouterCommand(
           throwIfCommandAborted(signal);
           const rows = await runtime.store.listTelemetry({ limit: parsed.limit });
           throwIfCommandAborted(signal);
-          ctx.ui.notify(formatHistoryMessage(rows), 'info');
+          ctx.ui.notify(
+            formatHistoryMessage(rows, { fleet: runtime.streamDeps.fleet }),
+            'info',
+          );
           return;
         }
 

package/.pi/extensions/smart-router/index.ts

@@ -38,6 +38,8 @@ import {
   formatHistoryMessage,
   formatStatusMessage,
   parseSmartRouterArgs,
+  resolveHistoryModelId,
+  qualifyModelIdForDisplay,
 } from './command-formatters.js';
 import { formatPricingStalenessLine, refreshPricingCatalog } from './pricing-lifecycle.js';
 import {
@@ -87,8 +89,10 @@ export {
   getSmartRouterArgumentCompletions,
   mapContextMessages,
   parseSmartRouterArgs,
+  qualifyModelIdForDisplay,
   refreshPricingCatalog,
   resolveDelegationOptions,
+  resolveHistoryModelId,
   logRoutingDecision,
   toDatasetExportRecord,
   capturePreRouteOutcomes,

package/README.md

@@ -197,7 +197,7 @@ Cursor models bill against your **Cursor Pro subscription quota**, not per-token
 |---------|---------|
 | `/smart-router` | Same as `status` (default when no subcommand is given) |
 | `/smart-router status` | Show fleet mode, fleet size, pricing freshness/staleness, and the last routing decision (stage, tier, selected model, latency) |
-| `/smart-router history` | Show recent routing telemetry from SQLite (default limit; optional numeric limit, e.g. `/smart-router history 20`) |
+| `/smart-router history` | Show recent routing telemetry from SQLite (default limit; optional numeric limit, e.g. `/smart-router history 20`). Displays the concrete delegated model id (never bare virtual `auto`) |
 | `/smart-router mode scoped` | Route only among pi's **enabled model patterns** (default) |
 | `/smart-router mode all` | Route among **all authenticated models** in the registry |
 | `/smart-router pricing refresh` | Manually fetch LiteLLM pricing from `LITELLM_PRICING_URL`, persist to SQLite, and rebuild the fleet with updated rates |
@@ -339,7 +339,7 @@ Cluster IDs are stable reason-code prefixes (`cluster_low_stakes_general`, `clus
 | Variable | Default | Purpose |
 |----------|---------|---------|
 | `ROUTER_STATE_DB_PATH` | `./.pi-smart-router/state.db` | Override SQLite state store location (telemetry, pricing catalog, session data) |
-| `SMART_ROUTER_LOG_ROUTING` | (unset) | Set to `1` to log each routing decision to stderr as JSON (debugging dogfood sessions) |
+| `SMART_ROUTER_LOG_ROUTING` | (unset) | Set to `1` to log each routing decision to stderr as JSON (debugging dogfood sessions). Canonical payload builder (`buildRoutingDecisionLogPayload`) includes top-level `stage`, `reason_code`, `low_intensity_score`, `tier_hint`, `local_eligible_reason`, and `cluster_id` (plus nested `cluster_summary` / `features`). The pi extension’s live stderr logger is still a slim subset — see [LOG_ROUTING field checklist](#log_routing-field-checklist) |
 | `SMART_ROUTER_DATASET` | (unset) | Set to `1` to opt in to privacy-safe routing dataset capture (metadata and feature fields only; 30-day / 10k-row retention). Prompt text, messages, and tool arguments are never stored. Required for outcome labels and P(success) training export. See [#8](https://github.com/beettlle/pi-smart-router/issues/8). |
 | `SMART_ROUTER_DATASET_FINGERPRINT` | (unset) | Set to `1` (requires `SMART_ROUTER_DATASET=1`) to store an install-local HMAC-SHA256 fingerprint of each normalized prompt for duplicate detection within this install. The install pepper lives in `.pi-smart-router/.dataset-key` (gitignored) and is never exported. **Warning:** short or common prompts are vulnerable to offline rainbow-table guessing; use only when you accept that tradeoff. See [#10](https://github.com/beettlle/pi-smart-router/issues/10). |
 | `MODELS_YAML_PATH` | `./config/models.yaml` | Fleet catalog path (library API only) |
@@ -354,6 +354,23 @@ Cluster IDs are stable reason-code prefixes (`cluster_low_stakes_general`, `clus
 | `ROUTER_SAFE_DEFAULT_TIER` | `economical-cloud` | Fallback tier on any routing failure |
 | `LITELLM_PRICING_URL` | — | LiteLLM pricing JSON source |
 
+### LOG_ROUTING field checklist
+
+When `SMART_ROUTER_LOG_ROUTING=1`, prefer the canonical payload from `buildRoutingDecisionLogPayload` (library / tests). Checklist for [#99](https://github.com/beettlle/pi-smart-router/issues/99):
+
+| Field | In payload builder? | Notes |
+|-------|---------------------|-------|
+| `stage` | Yes (top-level) | Pipeline stage that decided |
+| `reason_code` | Yes (top-level) | Machine-readable reason |
+| `low_intensity_score` | Yes (top-level + `cluster_summary`) | Null when low-intensity stage did not run |
+| `tier_hint` | Yes (top-level + `cluster_summary`) | Null when no tier hint |
+| `local_eligible_reason` | Yes (top-level + `features`) | Null when local_zero did not evaluate eligibility |
+| `cluster_id` | Yes (top-level + `cluster_summary`) | Null when no cluster match |
+
+**Gap:** the pi extension’s live stderr path (`logRoutingDecision` in `.pi/extensions/smart-router`) still emits a slim JSON object (`selected_model_id`, `stage`, `reason_code`, `features`, `delegate`) and does **not** yet call `buildRoutingDecisionLogPayload`. SQLite `/smart-router history` and the payload builder carry the full checklist; wire the extension logger in a follow-up if dogfood needs identical stderr shape.
+
+**History model id:** `/smart-router history` resolves bare/`smart-router` virtual `auto` to the concrete planning-delegate primary (or qualifies Cursor opaque `auto` as `cursor/auto`) so operators see the delegated fleet model, not the virtual router id.
+
 ### SAAR session pin and cache breakeven (v0.2.0 Continuity)
 
 v0.2.0 adds **Session-Aware Agentic Routing (SAAR)** pin knobs ([#72](https://github.com/beettlle/pi-smart-router/issues/72)) and a **cache breakeven gate** ([#73](https://github.com/beettlle/pi-smart-router/issues/73)) that blocks tier switches when `marginal_savings + future_cache_value <= cache_reprime_cost` — preventing cheap-turn savings from invalidating a warm prefix cache.
@@ -567,8 +584,10 @@ Additional operator defaults:
 
 | Key | Default | Purpose |
 |-----|---------|---------|
-| `loop_escalation.threshold` | 3 | Consecutive identical failures before escalating to frontier |
+| `loop_escalation.threshold` | 3 | Consecutive identical failures before escalating to frontier. Also used as the default **zero-tier tool-call churn** threshold (SP-178 / [#99](https://github.com/beettlle/pi-smart-router/issues/99)): while pinned to `zero-tier`, unsupported/unknown tool results escalate immediately, and N `tool_result` turns escalate via the same `loop_escalation` pin path (FR-014) — not a cache-breakeven bypass |
 | `pin_only_fallback` | `false` | Emergency pin-on-first-turn mode — see [Pin-only emergency fallback](#pin-only-emergency-fallback) |
+| `local_zero.enabled` | `true` | When `false`, skip `local_zero` dispatch (fall through to later stages). Default keeps the cheap local path for true trivial traffic |
+| `local_zero.max_tool_use_requirement` | `0.25` | Ceiling (0–1) on cheap predicted tool_use for `local_zero`. Effective limit is `min(local model tool_use, this value)`. Skips agentic git/bash/edit/explore/delete/repo cues with telemetry reason `tool_use_capability_shortfall` (SP-177 / [#98](https://github.com/beettlle/pi-smart-router/issues/98)) |
 | `local.min_memory_gb_full` | 16 | Minimum RAM for full local inference |
 | `local.battery_threshold_pct` | 20 | Minimum battery to allow local inference |
 | `pricing.staleness_days` | 14 | Max age before re-fetching pricing data |
@@ -809,7 +828,7 @@ npm run routing:eval-replay
 
 ### Benchmark profile refresh
 
-Capability scores in `config/benchmark-profiles.json` are grounded from public leaderboard snapshots under `tests/fixtures/benchmark-leaderboards/`. Each artifact records provenance (`source_urls`, `scrape_date`, `catalog_freeze_date`) in its header.
+Capability scores in `config/benchmark-profiles.json` are grounded from public leaderboard snapshots under `tests/fixtures/benchmark-leaderboards/` (and optional **recorded** live snapshots under `tests/fixtures/benchmark-leaderboards/recorded/`). Each artifact records provenance (`source_urls`, `scrape_date`, `catalog_freeze_date`) in its header.
 
 **Fleet ID aliases (SP-174):** live pi/Cursor scoped-fleet model IDs often differ from leaderboard `model_id` strings. The artifact’s optional `aliases` map sends those fleet IDs to an existing grounded row (never invents scores). `mapPiModelToProfile` sets `capability_source` to `benchmark` when a direct row or alias hits, otherwise `pattern_default`. Operators can also call `getCapabilitySource(modelId)` / `resolveBenchmarkModelId(modelId)`.
 
@@ -821,16 +840,61 @@ Capability scores in `config/benchmark-profiles.json` are grounded from public l
 4. Re-run ingest anytime — the CLI **preserves** existing `aliases` from the output file. Seed defaults live in `DEFAULT_FLEET_BENCHMARK_ALIASES` when no prior artifact exists.
 5. Confirm with `npm run routing:verify-benchmark-profiles` and a mapper unit test that `capability_source === 'benchmark'` for the fleet id.
 
+**Operator refresh command (SP-179 / SP-180):**
+
+| Mode | Command | Network? | When to use |
+|------|---------|----------|-------------|
+| **Fixtures (default)** | `npm run routing:ingest-benchmarks` | No | Local edits, CI, PR smoke |
+| **Recorded replay** | `npm run routing:ingest-benchmarks -- --recorded` | No | Replay last successful live snapshots offline |
+| **Live + record** | `npm run routing:ingest-benchmarks -- --live` | Yes | Operator refresh; writes `tests/fixtures/benchmark-leaderboards/recorded/` then regenerates profiles |
+
+Optional flags: `--catalog-freeze-date YYYY-MM-DD`, `--scrape-date YYYY-MM-DD`, `--record-dir DIR`, `--live-url BENCHMARK=URL`, `--output PATH`. See `npm run routing:ingest-benchmarks -- --help`.
+
+**Live sources (per benchmark):** each `--live` run resolves independently — live adapter → recorded → checked-in fixtures. One failing source never invents scores or blocks siblings. Logs: `ingest-benchmark-profiles: <id> source=live|recorded|fixture (…)`.
+
+| Benchmark | Default live fetch | Score field | Fallback |
+|-----------|-------------------|-------------|----------|
+| `swebench_verified` | Native: [SWE-bench `leaderboards.json`](https://raw.githubusercontent.com/SWE-bench/swe-bench.github.io/master/data/leaderboards.json) (Verified board) | `resolved` → `score` (0–100) | recorded → fixtures |
+| `livecodebench` | Native: [LCB `performances_generation.json`](https://raw.githubusercontent.com/LiveCodeBench/livecodebench.github.io/main/src/mocks/performances_generation.json) | aggregate `pass@1` | recorded → fixtures |
+| `bfcl` | Native: [Gorilla `data_overall.csv`](https://raw.githubusercontent.com/ShishirPatil/gorilla/gh-pages/data_overall.csv) | `Overall Acc` | recorded → fixtures |
+| `terminal_bench` | **No free stable JSON** (tbench.ai is HTML; HF leaderboard is submissions-only; paid Parse API is **not** the default). Pass `--live-url terminal_bench=URL` at a fixture-shaped mirror | `score` (0–100) | recorded → fixtures |
+
+**Terminal-Bench operator mirror schema** (SP-185 / #104):
+
+```json
+{
+  "benchmark": "terminal_bench",
+  "source_url": "https://www.tbench.ai/leaderboard",
+  "scrape_date": "YYYY-MM-DD",
+  "entries": [{ "model_id": "claude-opus-4-5", "score": 72.5 }]
+}
+```
+
+Example: `npm run routing:ingest-benchmarks -- --live --live-url terminal_bench=https://example.com/tb-mirror.json`. HTML bodies fail fast; without `--live-url`, TB uses recorded/fixtures. `release:refresh-benchmarks` uses the same `--live` path (fixture fallback on total failure).
+
+**Cadence (release-tied, not calendar):**
+
+| Trigger | When | Behavior |
+|---------|------|----------|
+| **Pre-tag release gate** | `npm run release:check` → `release:refresh-benchmarks` | Attempt **live** ingest; on failure fall back to fixtures; **fail if** `config/benchmark-profiles.json` or recorded snapshots are dirty — commit on `main`, re-run, then `npm version` / tag |
+| **Manual dispatch** | Actions → *Benchmark Profile Refresh* → `workflow_dispatch` (`use_live` default `true`) | Same live-or-fixture path; opens a bot PR when scores change; set `use_live=false` for fixtures-only |
+| **PR smoke** | PRs touching fixtures / ingest / artifact / workflow | **Fixtures only** — `npm run routing:verify-benchmark-profiles` (offline, no network) |
+
+There is **no monthly cron**. Refresh runs when you ship so each release packages the latest grounded scores. Tag-triggered Release publish uses `--ignore-scripts` and does not re-fetch (profiles are already frozen in the tag). Offline skip: `SMART_ROUTER_SKIP_LIVE_BENCHMARK_REFRESH=1 npm run release:check`.
+
 **Operator policy:**
 
-1. **PR smoke** — `.github/workflows/benchmark-profile-refresh.yml` runs on PRs that touch fixtures, ingest, or the checked-in artifact. It executes `npm run routing:verify-benchmark-profiles` so fixture edits cannot drift from `config/benchmark-profiles.json`.
-2. **Monthly refresh** — the same workflow runs on the 1st of each month (06:00 UTC) and via `workflow_dispatch`. It re-ingests fixtures, updates `catalog_freeze_date` to the run date, and opens a PR when model scores change.
-3. **Manual updates** — after editing fixture snapshots, run `npm run routing:ingest-benchmarks` (optionally `--catalog-freeze-date YYYY-MM-DD`) and commit the regenerated `config/benchmark-profiles.json` with the PR.
+1. **PR smoke** — fixture-only verify so PRs never require live network.
+2. **Every release** — live with fixture fallback via `release:check`; commit any profile/recorded diffs on `main` before tagging.
+3. **Ad-hoc** — Actions dispatch or local `--live` when refreshing between releases.
+4. **Manual local updates** — prefer fixtures or `--recorded` for offline work; use `--live` when refreshing from public leaderboard JSON endpoints.
 
-Regenerate locally:
+Verify after any regenerate:
 
 ```bash
 npm run routing:ingest-benchmarks
+# or: npm run routing:ingest-benchmarks -- --live
+# or: npm run routing:ingest-benchmarks -- --recorded
 npm run routing:verify-benchmark-profiles
 ```
 
@@ -844,7 +908,7 @@ Tag-triggered publish via GitHub Actions (requires `NPMSECRET` repository secret
 2. `routing:verify-benchmark-profiles` — checked-in capability profiles match fixture ingest
 3. `assert-release-gates --fixtures tests/eval/fixtures --baseline-version 0.6.0` — eval harness aggregate metrics vs `config/release-gates.json` and semver baseline regression vs `tests/eval/baselines/v0.6.0.json`
 
-`release:check` runs the full pre-release path: `verify:ci`, consumer pack verify, then Tier 0 functional smoke.
+`release:check` runs the full pre-release path: **live benchmark profile refresh** (fixture fallback; dirty-tree fail), then `verify:ci`, consumer pack verify, then Tier 0 functional smoke.
 
 **Baseline re-capture (post-tag):** after shipping a new semver (e.g. v0.7.0), freeze harness metrics for the next regression reference:
 
@@ -857,7 +921,8 @@ npm run routing:capture-baseline -- --version 0.7.0
 
 Commit the new baseline JSON and update `baseline_regression.reference_version` in `config/release-gates.json` plus the `--baseline-version` flag in `release:functional-smoke`. Re-run `npm run release:check` before tagging the next release.
 
-1. `npm run release:check` (CI parity + consumer pack + Tier 0 functional smoke)
+1. `npm run release:check` (live benchmark refresh → CI parity + consumer pack + Tier 0 functional smoke)
+   - If refresh rewrites profiles/recorded snapshots, **commit them on `main`** and re-run until clean
 2. `npm version 0.1.1` (creates commit + `v0.1.1` tag)
 3. `git push && git push --tags`
 4. Actions → **Release** runs pack smoke, consumer pack verify, Tier 0 functional smoke, `npm publish`, and creates a GitHub Release

package/config/benchmark-profiles.json

@@ -7,8 +7,8 @@
       "livecodebench": "https://livecodebench.github.io/leaderboard.html",
       "bfcl": "https://gorilla.cs.berkeley.edu/leaderboard.html"
     },
-    "scrape_date": "2026-07-09",
-    "catalog_freeze_date": "2026-07-09"
+    "scrape_date": "2026-07-11",
+    "catalog_freeze_date": "2026-07-11"
   },
   "aliases": {
     "claude-3-5-sonnet": "claude-sonnet-4-6",
@@ -34,50 +34,24 @@
   },
   "models": [
     {
-      "model_id": "claude-3.5-haiku",
+      "model_id": "claude-opus-4-5",
       "capabilities": {
-        "reasoning": 0.5685,
-        "code_gen": 0.587,
-        "tool_use": 0.6095
+        "reasoning": 0.7585,
+        "code_gen": 0.708,
+        "tool_use": 0.7498
       },
       "benchmark_sources": {
         "bfcl": {
-          "raw_score": 70.6,
-          "normalized": 0.706
+          "raw_score": 77.47,
+          "normalized": 0.7746999999999999
         },
         "livecodebench": {
-          "raw_score": 55,
-          "normalized": 0.55
-        },
-        "swebench_verified": {
           "raw_score": 62.4,
           "normalized": 0.624
         },
-        "terminal_bench": {
-          "raw_score": 51.3,
-          "normalized": 0.513
-        }
-      }
-    },
-    {
-      "model_id": "claude-opus-4-5",
-      "capabilities": {
-        "reasoning": 0.767,
-        "code_gen": 0.7915,
-        "tool_use": 0.8035
-      },
-      "benchmark_sources": {
-        "bfcl": {
-          "raw_score": 88.2,
-          "normalized": 0.882
-        },
-        "livecodebench": {
-          "raw_score": 77.4,
-          "normalized": 0.774
-        },
         "swebench_verified": {
-          "raw_score": 80.9,
-          "normalized": 0.809
+          "raw_score": 79.2,
+          "normalized": 0.792
         },
         "terminal_bench": {
           "raw_score": 72.5,
@@ -88,22 +62,22 @@
     {
       "model_id": "claude-sonnet-4-6",
       "capabilities": {
-        "reasoning": 0.738,
-        "code_gen": 0.7635,
-        "tool_use": 0.7625
+        "reasoning": 0.724,
+        "code_gen": 0.681,
+        "tool_use": 0.7062
       },
       "benchmark_sources": {
         "bfcl": {
-          "raw_score": 84.5,
-          "normalized": 0.845
+          "raw_score": 73.24,
+          "normalized": 0.7323999999999999
         },
         "livecodebench": {
-          "raw_score": 73.1,
-          "normalized": 0.731
+          "raw_score": 59.4,
+          "normalized": 0.594
         },
         "swebench_verified": {
-          "raw_score": 79.6,
-          "normalized": 0.7959999999999999
+          "raw_score": 76.8,
+          "normalized": 0.768
         },
         "terminal_bench": {
           "raw_score": 68,
@@ -114,22 +88,22 @@
     {
       "model_id": "gemini-2.5-flash",
       "capabilities": {
-        "reasoning": 0.5295,
-        "code_gen": 0.5525,
-        "tool_use": 0.5735
+        "reasoning": 0.3827,
+        "code_gen": 0.5192,
+        "tool_use": 0.5202
       },
       "benchmark_sources": {
         "bfcl": {
-          "raw_score": 66.9,
-          "normalized": 0.669
+          "raw_score": 56.24,
+          "normalized": 0.5624
         },
         "livecodebench": {
-          "raw_score": 52.4,
-          "normalized": 0.524
+          "raw_score": 75.1,
+          "normalized": 0.7509999999999999
         },
         "swebench_verified": {
-          "raw_score": 58.1,
-          "normalized": 0.581
+          "raw_score": 28.73,
+          "normalized": 0.2873
         },
         "terminal_bench": {
           "raw_score": 47.8,
@@ -140,22 +114,14 @@
     {
       "model_id": "gpt-5.3-codex",
       "capabilities": {
-        "reasoning": 0.796,
-        "code_gen": 0.833,
-        "tool_use": 0.8215
+        "reasoning": 0.743,
+        "code_gen": 0.744,
+        "tool_use": 0.742
       },
       "benchmark_sources": {
-        "bfcl": {
-          "raw_score": 90.1,
-          "normalized": 0.9009999999999999
-        },
-        "livecodebench": {
-          "raw_score": 81.6,
-          "normalized": 0.816
-        },
         "swebench_verified": {
-          "raw_score": 85,
-          "normalized": 0.85
+          "raw_score": 74.4,
+          "normalized": 0.7440000000000001
         },
         "terminal_bench": {
           "raw_score": 74.2,

package/config/operator-config.json.example

@@ -60,5 +60,13 @@
       "exclude_execution_history": true
     }
   },
+  "local_zero": {
+    "enabled": true,
+    "max_tool_use_requirement": 0.25,
+    "_documentation": {
+      "enabled": "When false, skip local_zero entirely (fall through to triage cloud / HyDRA). Default true — does not disable local for trivial traffic.",
+      "max_tool_use_requirement": "Ceiling (0–1) on cheap predicted tool_use for local_zero dispatch. Effective limit is min(local model tool_use capability, this value). Default 0.25 keeps format/lint local while skipping agentic git/bash/edit/explore/delete/repo cues (SP-177, #98)."
+    }
+  },
   "pin_only_fallback": false
 }

package/dist/config/defaults.d.ts

@@ -3,7 +3,7 @@
  * Values sourced from specs/001-build-smart-router/data-model.md § Configuration (Operator).
  */
 import { type OperatorConfig } from '../domain/types/schemas.js';
-export { DEFAULT_PLANNING_DELEGATE_CONFIG, DEFAULT_SAAR_CONFIG, resolvePlanningDelegateConfigFromEnv, resolveSaarConfigFromEnv } from '../domain/types/schemas.js';
+export { DEFAULT_LOCAL_ZERO_CONFIG, DEFAULT_PLANNING_DELEGATE_CONFIG, DEFAULT_SAAR_CONFIG, resolvePlanningDelegateConfigFromEnv, resolveSaarConfigFromEnv, } from '../domain/types/schemas.js';
 /** Merge operator env overrides onto defaults (SAAR and planning delegate sections). */
 export declare function resolveOperatorConfigFromEnv(base?: OperatorConfig): OperatorConfig;
 export declare const DEFAULT_OPERATOR_CONFIG: Readonly<OperatorConfig>;

package/dist/config/defaults.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"defaults.d.ts","sourceRoot":"","sources":["../../src/config/defaults.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,EAKL,KAAK,cAAc,EACpB,MAAM,4BAA4B,CAAC;AAGpC,OAAO,EAAE,gCAAgC,EAAE,mBAAmB,EAAE,oCAAoC,EAAE,wBAAwB,EAAE,MAAM,4BAA4B,CAAC;AAEnK,wFAAwF;AACxF,wBAAgB,4BAA4B,CAC1C,IAAI,GAAE,cAAwC,GAC7C,cAAc,CAMhB;AAED,eAAO,MAAM,uBAAuB,EAAE,QAAQ,CAAC,cAAc,CA+BnD,CAAC"}
+{"version":3,"file":"defaults.d.ts","sourceRoot":"","sources":["../../src/config/defaults.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,EAML,KAAK,cAAc,EACpB,MAAM,4BAA4B,CAAC;AAGpC,OAAO,EACL,yBAAyB,EACzB,gCAAgC,EAChC,mBAAmB,EACnB,oCAAoC,EACpC,wBAAwB,GACzB,MAAM,4BAA4B,CAAC;AAEpC,wFAAwF;AACxF,wBAAgB,4BAA4B,CAC1C,IAAI,GAAE,cAAwC,GAC7C,cAAc,CAMhB;AAED,eAAO,MAAM,uBAAuB,EAAE,QAAQ,CAAC,cAAc,CAgCnD,CAAC"}

package/dist/config/defaults.js

@@ -2,9 +2,9 @@
  * Operator configuration defaults (FR-021).
  * Values sourced from specs/001-build-smart-router/data-model.md § Configuration (Operator).
  */
-import { DEFAULT_PLANNING_DELEGATE_CONFIG, DEFAULT_SAAR_CONFIG, resolvePlanningDelegateConfigFromEnv, resolveSaarConfigFromEnv, } from '../domain/types/schemas.js';
+import { DEFAULT_LOCAL_ZERO_CONFIG, DEFAULT_PLANNING_DELEGATE_CONFIG, DEFAULT_SAAR_CONFIG, resolvePlanningDelegateConfigFromEnv, resolveSaarConfigFromEnv, } from '../domain/types/schemas.js';
 import { DEFAULT_LOW_INTENSITY_WEIGHTS } from '../domain/routing/tier-features.js';
-export { DEFAULT_PLANNING_DELEGATE_CONFIG, DEFAULT_SAAR_CONFIG, resolvePlanningDelegateConfigFromEnv, resolveSaarConfigFromEnv } from '../domain/types/schemas.js';
+export { DEFAULT_LOCAL_ZERO_CONFIG, DEFAULT_PLANNING_DELEGATE_CONFIG, DEFAULT_SAAR_CONFIG, resolvePlanningDelegateConfigFromEnv, resolveSaarConfigFromEnv, } from '../domain/types/schemas.js';
 /** Merge operator env overrides onto defaults (SAAR and planning delegate sections). */
 export function resolveOperatorConfigFromEnv(base = DEFAULT_OPERATOR_CONFIG) {
     return {
@@ -43,6 +43,7 @@ export const DEFAULT_OPERATOR_CONFIG = {
     },
     saar: DEFAULT_SAAR_CONFIG,
     planning_delegate: DEFAULT_PLANNING_DELEGATE_CONFIG,
+    local_zero: DEFAULT_LOCAL_ZERO_CONFIG,
     pin_only_fallback: false,
 };
 //# sourceMappingURL=defaults.js.map

package/dist/config/defaults.js.map

@@ -1 +1 @@
-{"version":3,"file":"defaults.js","sourceRoot":"","sources":["../../src/config/defaults.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,EACL,gCAAgC,EAChC,mBAAmB,EACnB,oCAAoC,EACpC,wBAAwB,GAEzB,MAAM,4BAA4B,CAAC;AACpC,OAAO,EAAE,6BAA6B,EAAE,MAAM,oCAAoC,CAAC;AAEnF,OAAO,EAAE,gCAAgC,EAAE,mBAAmB,EAAE,oCAAoC,EAAE,wBAAwB,EAAE,MAAM,4BAA4B,CAAC;AAEnK,wFAAwF;AACxF,MAAM,UAAU,4BAA4B,CAC1C,OAAuB,uBAAuB;IAE9C,OAAO;QACL,GAAG,IAAI;QACP,IAAI,EAAE,wBAAwB,CAAC,IAAI,CAAC,IAAI,CAAC;QACzC,iBAAiB,EAAE,oCAAoC,CAAC,IAAI,CAAC,iBAAiB,CAAC;KAChF,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,MAAM,uBAAuB,GAA6B;IAC/D,SAAS,EAAE;QACT,WAAW,EAAE,GAAG;QAChB,cAAc,EAAE,GAAG;QACnB,gBAAgB,EAAE,IAAI;KACvB;IACD,eAAe,EAAE;QACf,SAAS,EAAE,CAAC;KACb;IACD,OAAO,EAAE;QACP,cAAc,EAAE,EAAE;KACnB;IACD,KAAK,EAAE;QACL,kBAAkB,EAAE,EAAE;QACtB,4BAA4B,EAAE,CAAC;QAC/B,qBAAqB,EAAE,EAAE;KAC1B;IACD,KAAK,EAAE;QACL,mBAAmB,EAAE,0BAA0B;QAC/C,OAAO,EAAE,QAAQ;QACjB,WAAW,EAAE,oBAAoB;KAClC;IACD,aAAa,EAAE;QACb,OAAO,EAAE,6BAA6B;QACtC,cAAc,EAAE,IAAI;QACpB,aAAa,EAAE,IAAI;QACnB,eAAe,EAAE,GAAG;KACrB;IACD,IAAI,EAAE,mBAAmB;IACzB,iBAAiB,EAAE,gCAAgC;IACnD,iBAAiB,EAAE,KAAK;CAChB,CAAC"}
+{"version":3,"file":"defaults.js","sourceRoot":"","sources":["../../src/config/defaults.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,EACL,yBAAyB,EACzB,gCAAgC,EAChC,mBAAmB,EACnB,oCAAoC,EACpC,wBAAwB,GAEzB,MAAM,4BAA4B,CAAC;AACpC,OAAO,EAAE,6BAA6B,EAAE,MAAM,oCAAoC,CAAC;AAEnF,OAAO,EACL,yBAAyB,EACzB,gCAAgC,EAChC,mBAAmB,EACnB,oCAAoC,EACpC,wBAAwB,GACzB,MAAM,4BAA4B,CAAC;AAEpC,wFAAwF;AACxF,MAAM,UAAU,4BAA4B,CAC1C,OAAuB,uBAAuB;IAE9C,OAAO;QACL,GAAG,IAAI;QACP,IAAI,EAAE,wBAAwB,CAAC,IAAI,CAAC,IAAI,CAAC;QACzC,iBAAiB,EAAE,oCAAoC,CAAC,IAAI,CAAC,iBAAiB,CAAC;KAChF,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,MAAM,uBAAuB,GAA6B;IAC/D,SAAS,EAAE;QACT,WAAW,EAAE,GAAG;QAChB,cAAc,EAAE,GAAG;QACnB,gBAAgB,EAAE,IAAI;KACvB;IACD,eAAe,EAAE;QACf,SAAS,EAAE,CAAC;KACb;IACD,OAAO,EAAE;QACP,cAAc,EAAE,EAAE;KACnB;IACD,KAAK,EAAE;QACL,kBAAkB,EAAE,EAAE;QACtB,4BAA4B,EAAE,CAAC;QAC/B,qBAAqB,EAAE,EAAE;KAC1B;IACD,KAAK,EAAE;QACL,mBAAmB,EAAE,0BAA0B;QAC/C,OAAO,EAAE,QAAQ;QACjB,WAAW,EAAE,oBAAoB;KAClC;IACD,aAAa,EAAE;QACb,OAAO,EAAE,6BAA6B;QACtC,cAAc,EAAE,IAAI;QACpB,aAAa,EAAE,IAAI;QACnB,eAAe,EAAE,GAAG;KACrB;IACD,IAAI,EAAE,mBAAmB;IACzB,iBAAiB,EAAE,gCAAgC;IACnD,UAAU,EAAE,yBAAyB;IACrC,iBAAiB,EAAE,KAAK;CAChB,CAAC"}

package/dist/domain/pinning/loop-escalation.d.ts

@@ -1,5 +1,5 @@
 /**
- * Loop escalation — FR-014, FR-008 rule #3.
+ * Loop escalation — FR-014, FR-008 rule #3 / #4.
  *
  * Detects bounded repeated identical tool failures and signals
  * session pin escalation to a frontier-capable tier.
@@ -7,12 +7,27 @@
  * Escalation fires once per session; no tier oscillation (FR-008).
  * Threshold defaults to 3 identical tool failures (operator configurable).
  *
+ * Zero-tier observational pin-break (SP-178 / #99):
+ * SAAR pins preserve prefix-cache value, but a zero-tier pin that faces
+ * unsupported/unknown tools or sustained tool-loop churn is a capability
+ * mismatch — the same class of observational signal as identical tool
+ * failures (FR-014). Escalation reuses the `loop_escalation` pin reason
+ * (allowed FR-008 break) rather than inventing a cache-economics bypass.
+ * Voluntary cross-provider switches still go through breakeven; this path
+ * only fires on observational evidence that the pinned zero-tier model
+ * cannot complete the tool loop.
+ *
  * Design: pure evaluation function — caller (pipeline stage) is
  * responsible for applying pin state updates via SessionPinner.
  */
 import type { ModelProfile, RoutingRequest, SessionPin } from '../types/index.js';
 export interface LoopEscalationConfig {
     readonly threshold: number;
+    /**
+     * Tool-result turns while pinned to zero-tier before observational escalate.
+     * Defaults to `threshold` when omitted (same operator knob, no schema change).
+     */
+    readonly zero_tier_tool_call_threshold?: number;
 }
 export interface LoopEscalationResult {
     readonly shouldEscalate: boolean;
@@ -20,6 +35,8 @@ export interface LoopEscalationResult {
     readonly escalationTarget: ModelProfile | null;
     readonly reason: string;
 }
+/** Stable signature for zero-tier tool-call churn counting (SP-178). */
+export declare const ZERO_TIER_TOOL_CHURN_SIGNATURE: "zt:tool_churn";
 /**
  * Extract a tool-failure signature from the request's messages.
  * Returns null when the request does not carry a tool failure.
@@ -28,6 +45,11 @@ export interface LoopEscalationResult {
  * carry observational failure signals (FR-014: no post-generation judging).
  */
 export declare function extractToolFailureSignature(request: RoutingRequest): string | null;
+/**
+ * True when the latest tool result reports an unsupported/unknown tool
+ * (capability mismatch — escalate immediately on zero-tier pins).
+ */
+export declare function isUnsupportedOrUnknownToolResult(request: RoutingRequest): boolean;
 /**
  * Evaluate whether the session should escalate to a higher tier.
  *

package/dist/domain/pinning/loop-escalation.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"loop-escalation.d.ts","sourceRoot":"","sources":["../../../src/domain/pinning/loop-escalation.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;GAWG;AAEH,OAAO,KAAK,EAAE,YAAY,EAAE,cAAc,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAIlF,MAAM,WAAW,oBAAoB;IACnC,QAAQ,CAAC,SAAS,EAAE,MAAM,CAAC;CAC5B;AAED,MAAM,WAAW,oBAAoB;IACnC,QAAQ,CAAC,cAAc,EAAE,OAAO,CAAC;IACjC,QAAQ,CAAC,UAAU,EAAE,UAAU,GAAG,IAAI,CAAC;IACvC,QAAQ,CAAC,gBAAgB,EAAE,YAAY,GAAG,IAAI,CAAC;IAC/C,QAAQ,CAAC,MAAM,EAAE,MAAM,CAAC;CACzB;AAgBD;;;;;;GAMG;AACH,wBAAgB,2BAA2B,CAAC,OAAO,EAAE,cAAc,GAAG,MAAM,GAAG,IAAI,CAclF;AAmBD;;;;;;;;GAQG;AACH,wBAAgB,sBAAsB,CACpC,GAAG,EAAE,UAAU,GAAG,IAAI,EACtB,OAAO,EAAE,cAAc,EACvB,KAAK,EAAE,SAAS,YAAY,EAAE,EAC9B,MAAM,EAAE,oBAAoB,GAC3B,oBAAoB,CAkEtB"}
+{"version":3,"file":"loop-escalation.d.ts","sourceRoot":"","sources":["../../../src/domain/pinning/loop-escalation.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;GAqBG;AAEH,OAAO,KAAK,EAAE,YAAY,EAAE,cAAc,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAIlF,MAAM,WAAW,oBAAoB;IACnC,QAAQ,CAAC,SAAS,EAAE,MAAM,CAAC;IAC3B;;;OAGG;IACH,QAAQ,CAAC,6BAA6B,CAAC,EAAE,MAAM,CAAC;CACjD;AAED,MAAM,WAAW,oBAAoB;IACnC,QAAQ,CAAC,cAAc,EAAE,OAAO,CAAC;IACjC,QAAQ,CAAC,UAAU,EAAE,UAAU,GAAG,IAAI,CAAC;IACvC,QAAQ,CAAC,gBAAgB,EAAE,YAAY,GAAG,IAAI,CAAC;IAC/C,QAAQ,CAAC,MAAM,EAAE,MAAM,CAAC;CACzB;AAED,wEAAwE;AACxE,eAAO,MAAM,8BAA8B,EAAG,eAAwB,CAAC;AA4BvE;;;;;;GAMG;AACH,wBAAgB,2BAA2B,CAAC,OAAO,EAAE,cAAc,GAAG,MAAM,GAAG,IAAI,CAOlF;AAED;;;GAGG;AACH,wBAAgB,gCAAgC,CAAC,OAAO,EAAE,cAAc,GAAG,OAAO,CAKjF;AA+GD;;;;;;;;GAQG;AACH,wBAAgB,sBAAsB,CACpC,GAAG,EAAE,UAAU,GAAG,IAAI,EACtB,OAAO,EAAE,cAAc,EACvB,KAAK,EAAE,SAAS,YAAY,EAAE,EAC9B,MAAM,EAAE,oBAAoB,GAC3B,oBAAoB,CAwDtB"}