@deeeed/metamask-harness 0.54.0 → 0.55.0

package/CHANGELOG.md

@@ -2,6 +2,23 @@
 
 ## Unreleased
 
+## 0.55.0 - 2026-09-18
+
+### Added
+
+- `doctor` reports two more setup answers, also carried in its `--json`: `update` says whether this harness is the published latest (`0.54.1 is latest`, `0.54.1 → 0.55.0 available`, or `not checked` when the registry is unreachable or `MM_HARNESS_NO_UPDATE_CHECK=1`), and `skills` counts the `mms-*` skill directories the checkout has installed (`none installed` when there are none). Both are advisory — neither joins `requiredChecks`, changes `status`/`ready`, or can fail doctor, because the harness does its job without either. The registry check answers from the existing update cache, so it costs nothing on a checkout that already looked today. A setup skill now reads one verdict from `doctor` instead of running checks of its own; which skill names are expected or retired stays with the installer and that skill, so the harness carries no list of them.
+
+### Changed
+
+- The checkout view's fixture block is labelled `last fixtures set:`, because it reports the result of the last `fixtures set` this checkout ran, not the state of the wallet fixture file — which `doctor` already reports as `fixture.status`.
+- The mobile deps note — raised when the manifests are newer than the install markers and no baseline is recorded — says what was observed, why it is treated as fine, and the one case that calls for action, with the install command it already teaches for a blocked preflight. It is set off from the command's own output by a blank line and a `mm-harness note (deps):` label, so a reader can tell the harness is talking.
+
+## 0.54.1 - 2026-09-18
+
+### Changed
+
+- `help --json` quality section carries the verdict rules (pass, warn, fail), the shape of an independent critique, and per-adapter review notes for Mobile, Extension and Core; the separate recipe-quality skill is retired in favour of this guidance.
+
 ## 0.54.0 - 2026-09-17
 
 ### Changed

package/README.md

@@ -271,7 +271,7 @@ Unset `MM_HARNESS_BIN` to return to the published installation.
 
 ### Workflow entry points and completion
 
-Explicitly invoke `mms-recipe-dev`, `mms-recipe-fixbug`, `mms-recipe-qa` or the team's static review skill. `mms-recipe-cook` selects or resumes their shared Markdown checklist. Matching descriptions or changed paths never activates these workflows.
+Explicitly invoke `mms-recipe-dev`, `mms-recipe-fix-bug`, `mms-recipe-qa` or the team's static review skill. `mms-recipe-cook` selects or resumes their shared Markdown checklist. Matching descriptions or changed paths never activates these workflows.
 
 Static Perps review uses a generated source checklist without installing or launching the app harness. `recipe-qa` freezes a PR, range or daily change window and requires runtime smoke plus selected recipe proof. Official release validation uses its named release profile and proof lane.
 

package/dist/adapters/mobile/runtime-decision.js

@@ -11,6 +11,7 @@ import {
   supersededErrorCapture
 } from "@farmslot/recipe-harness/runtime/log-analysis";
 import { probeMetroPackager } from "@farmslot/recipe-harness/runtime/metro-probe";
+import { color } from "../../cli-color.js";
 import { recipeRuntimePath } from "../../paths.js";
 import { resolveMobileToolPath } from "../../../library/actions/mobile/platform/tool-paths.mjs";
 import { mobileProductMarkers } from "./deps-markers.js";
@@ -147,14 +148,16 @@ const launchActions = (target, clearMetro = false) => {
   actions.push({ id: "wait-for-bridge", cwd: target });
   return actions;
 };
+function depsInstallCommand(resolved) {
+  return fs.existsSync(path.join(resolved, "node_modules")) ? "`yarn setup:expo --no-build-ios --no-build-android`" : "`yarn install --immutable && yarn setup:expo --no-build-ios --no-build-android`";
+}
 async function decideMobileReadiness(target, options = {}) {
   const resolved = path.resolve(target);
   if (options.record) recordDepsBaseline(resolved);
   const fast = options.preflightMode === "fast";
   const report = await computeMobileReadiness(resolved, options, fast);
   if (fast && report.decision === "install") {
-    const hasNodeModules = fs.existsSync(path.join(resolved, "node_modules"));
-    const standalone = hasNodeModules ? "`yarn setup:expo --no-build-ios --no-build-android`" : "`yarn install --immutable && yarn setup:expo --no-build-ios --no-build-android`";
+    const standalone = depsInstallCommand(resolved);
     return {
       ...report,
       decision: "blocked",
@@ -184,6 +187,15 @@ async function decideMobileReadiness(target, options = {}) {
     actions: report.decision === "ready" ? launchActions(resolved) : report.actions
   };
 }
+function writeDepsNote(resolved) {
+  const stream = process.stderr;
+  const body = `package.json/yarn.lock are newer than the node_modules install markers and no install baseline is recorded yet, so the installed node_modules are used as they are. Pulling commits refreshes those timestamps without changing dependencies, and a required-dependency presence check still runs, so this on its own is not evidence that anything drifted. Act only if you know dependencies changed: run ${depsInstallCommand(resolved)} once \u2014 that refreshes the markers, and \`mm-harness prepare\` records a baseline when it runs the install, so this note stops.`;
+  stream.write(
+    `
+${color("warn", "mm-harness note (deps):", { stream })} ${color("dim", body, { stream })}
+`
+  );
+}
 async function computeMobileReadiness(resolved, options, fast) {
   const rawDeps = depsCheck(resolved, {
     productMarkers: mobileProductMarkers(options.platform)
@@ -191,9 +203,7 @@ async function computeMobileReadiness(resolved, options, fast) {
   let deps = rawDeps;
   if (rawDeps.status === "stale" && !rawDeps.hasBaseline) {
     deps = { installed: rawDeps.installed, status: "current", hasBaseline: false };
-    process.stderr.write(
-      "[runtime-decision] manifest is newer than install markers (no recorded baseline); trusting installed node_modules \u2014 run yarn setup manually if deps truly changed.\n"
-    );
+    writeDepsNote(resolved);
   }
   if (deps.status === "current") {
     const absent = missingRequiredDeps(resolved);

package/dist/checkout-view.js

@@ -47,7 +47,7 @@ function renderCheckoutView(view) {
     }
   }
   if (view.fixtures) {
-    lines.push(`${out("label", "fixtures:")}`);
+    lines.push(`${out("label", "last fixtures set:")}`);
     for (const [key, value] of Object.entries(view.fixtures)) {
       const isSet = value !== null && value !== "";
       lines.push(`  ${key.padEnd(22)} ${isSet ? out("ok", String(value)) : out("dim", "(unset)")}`);

package/dist/commands/doctor.js

@@ -21,6 +21,7 @@ import {
   renderMobileLiveBlock
 } from "./mobile-device-view.js";
 import { farmslotReadyIndicator } from "./farmslot-ready.js";
+import { updateStatus } from "./update.js";
 import { ensureRuntimeContext } from "../runtime-context.js";
 import { acquireCheckoutLock } from "../checkout-lock.js";
 import { ADAPTER_DETECT_NEXT, checkoutBusyOut, EXIT, usageOut } from "./shared.js";
@@ -206,12 +207,14 @@ async function handleDoctor({ options }) {
       discovery
     );
   }
+  const skills = skillsView(target);
+  const update = updateStatus();
   const doctorUserAction = doctorResult.checks.find(
     (check) => check.required && check.status === "fail" && check.userAction
   )?.userAction ?? `mm-harness doctor --fix --adapter ${adapter} --target ${shellQuote(target)} --json`;
   const doctorError = doctorResult.status === "fail" ? { code: "DOCTOR_CHECKS_FAILED", message: "one or more required doctor checks failed", userAction: doctorUserAction } : void 0;
   const next = doctorError?.userAction ?? discovery.deviceDiscoveryErrors?.[0]?.userAction ?? (runtime?.decision === "ready" ? void 0 : runtime?.nextAction);
-  if (json) console.log(JSON.stringify({ ...doctorResult, ready: runtime?.decision === "ready", runtime, view, orphanMetros, capture, devices, additionalReachableDevices, ...discovery, ...next ? { next } : {}, ...doctorError ? { error: doctorError } : {} }, null, 2));
+  if (json) console.log(JSON.stringify({ ...doctorResult, ready: runtime?.decision === "ready", runtime, view, orphanMetros, capture, skills, update, devices, additionalReachableDevices, ...discovery, ...next ? { next } : {}, ...doctorError ? { error: doctorError } : {} }, null, 2));
   else {
     const out = (style, text) => color(style, text, { stream: process.stdout });
     const stateStyle = (value, good) => value === good ? "ok" : "warn";
@@ -246,6 +249,8 @@ async function handleDoctor({ options }) {
       console.log(`  ${out("dim", "Next: mm-harness stop   # reaps every bundler this checkout leaked")}`);
     }
     if (capture) renderCaptureHealth(capture, out);
+    renderSkillsView(skills, out);
+    renderUpdateStatus(update, out);
     if (adapter === "mobile") {
       if (deviceView) renderMobileDeviceList(deviceView, out);
       if (additionalReachableDevices.length > 0) renderAdditionalReachableDevices(additionalReachableDevices, out);
@@ -374,6 +379,41 @@ function renderCaptureHealth(capture, out) {
     console.log(`  ${out("dim", "Next: screenshots work without capture-helper; install/configure capture-helper only when recipe video is required")}`);
   }
 }
+const SKILL_ROOTS = [path.join(".claude", "skills"), path.join(".agents", "skills")];
+function skillsView(target) {
+  const names = /* @__PURE__ */ new Set();
+  let dir = null;
+  for (const root of SKILL_ROOTS) {
+    const absolute = path.join(target, root);
+    let entries;
+    try {
+      entries = fs.readdirSync(absolute, { withFileTypes: true });
+    } catch {
+      continue;
+    }
+    dir ??= root;
+    for (const entry of entries) {
+      if (entry.isDirectory() && entry.name.startsWith("mms-")) names.add(entry.name);
+    }
+  }
+  return { dir, installed: [...names].sort() };
+}
+function renderSkillsView(skills, out) {
+  if (skills.installed.length === 0) {
+    console.log(`${out("label", "skills:")} ${out("warn", "none installed")} ${out("dim", "(yarn skills)")}`);
+    return;
+  }
+  console.log(`${out("label", "skills:")} ${skills.installed.length} installed ${out("dim", skills.dir ?? "")}`);
+}
+function renderUpdateStatus(status, out) {
+  if (!status.checked) {
+    console.log(`${out("label", "update:")} ${out("dim", "not checked")}`);
+    return;
+  }
+  console.log(
+    status.updateAvailable ? `${out("label", "update:")} ${out("warn", `${status.current} \u2192 ${status.latest} available`)} ${out("dim", "(mm-harness update)")}` : status.latest && status.latest !== status.current ? `${out("label", "update:")} ${out("ok", status.current)} ${out("dim", `ahead of registry ${status.latest}`)}` : `${out("label", "update:")} ${out("ok", `${status.current} is latest`)}`
+  );
+}
 function detectOrphanMetros(target, livePort) {
   try {
     const script = path.join(runnerDir, "adapters/shared/reap-checkout-metros.sh");
@@ -493,5 +533,6 @@ function readTextSnapshot(file) {
 export {
   doctorFixNextActions,
   ensureExtensionPerpsBuildFlag,
-  handleDoctor
+  handleDoctor,
+  skillsView
 };

package/dist/commands/update.js

@@ -199,6 +199,23 @@ function nudgeDisabled() {
 function nudgeLine(current, latest) {
   return isNewer(latest, current) ? `mm-harness ${current} \u2192 ${latest} available \xB7 run: mm-harness update` : null;
 }
+const DOCTOR_FETCH_TIMEOUT_MS = 3e3;
+const DOCTOR_CACHE_FRESH_MS = 60 * 60 * 1e3;
+function updateStatus(now = Date.now()) {
+  const current = currentVersion();
+  if (nudgeDisabled()) return { current, checked: false };
+  const file = cacheFile();
+  const cache = readCache(file);
+  const latest = cache?.latest && now - cache.lastCheck < DOCTOR_CACHE_FRESH_MS ? cache.latest : probeLatest(file, now);
+  if (!latest) return { current, checked: false };
+  return { current, checked: true, latest, updateAvailable: isNewer(latest, current) };
+}
+function probeLatest(file, now) {
+  const fetched = fetchLatest(DOCTOR_FETCH_TIMEOUT_MS);
+  if (!fetched.latest) return null;
+  writeCache(file, { lastCheck: now, latest: fetched.latest });
+  return fetched.latest;
+}
 function maybeNudge(now = Date.now()) {
   if (nudgeDisabled()) return;
   const file = cacheFile();
@@ -233,5 +250,6 @@ export {
   maybeNudge,
   nudgeDisabled,
   nudgeLine,
-  registryCwd
+  registryCwd,
+  updateStatus
 };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@deeeed/metamask-harness",
-  "version": "0.54.0",
+  "version": "0.55.0",
   "type": "module",
   "bin": {
     "mm-harness": "bin/mm-harness"

package/scripts/site-contrast.mjs

@@ -369,7 +369,7 @@ async function checkBehavior(browserEndpoint, port, pages) {
           await new Promise((resolve) => setTimeout(resolve, 25));
           return window.__siteCopied;
         }
-        const out = { setup: await grab(document.querySelector('.cmd-hero .copy, .prompt .copy')), work: [] };
+        const out = { setup: await grab(document.querySelector('.cmd-hero .copy, .prompt .copy')), work: [], text: document.body.innerText };
         for (const tab of document.querySelectorAll('[role="tab"]')) {
           tab.click();
           const panel = document.getElementById(tab.getAttribute('aria-controls'));
@@ -389,14 +389,15 @@ async function checkBehavior(browserEndpoint, port, pages) {
     // Setting up installs the harness and the skills; the work prompts invoke
     // them by their exact installed names.
     const expected = [
-      ['npm i -g @deeeed/metamask-harness@latest', copied.setup],
-      ['yarn skills', copied.setup],
+      ['/mms-recipe-setup', copied.setup],
+      ['agentic/recipe-setup', copied.text],
       ['/mms-recipe-qa', copied.work.join('\n')],
-      ['/mms-recipe-cook', copied.work.join('\n')],
+      ['/mms-recipe-fix-bug', copied.work.join('\n')],
+      ['/mms-recipe-dev', copied.work.join('\n')],
       ['/mms-recipe-performance', copied.work.join('\n')],
-      ['/mms-recipe-evidence', copied.work.join('\n')],
-      ['agentic/recipe-performance', copied.setup],
-      ['agentic/recipe-evidence', copied.setup],
+      ['/mms-recipe-pretty-pr', copied.work.join('\n')],
+      ['perps/recipe-performance', copied.text],
+      ['agentic/recipe-pretty-pr', copied.text],
     ];
     const missing = expected.filter(([line, text]) => !text?.includes(line)).map(([line]) => line);
     if (missing.length) {
@@ -448,7 +449,7 @@ async function checkBehavior(browserEndpoint, port, pages) {
         return checked;
       })()
     `);
-    if (!previews.includes('review-pr/static-perps.mobile') || !previews.includes('qa/pr.extension')) {
+    if (!previews.includes('review-pr/static-perps.mobile') || !previews.includes('qa/default')) {
       throw new Error('Static review and QA previews must both be available');
     }
     const progress = await page.evaluate(String.raw`

package/site/assets/help-recipes.json

@@ -212,6 +212,26 @@
             "mobile",
             "extension"
           ]
+        },
+        "Verdict: pass when every named claim ran an executable path with an assertion and consistent evidence; warn when the supported claims pass but non-blocking gaps remain; fail when a named claim has no executable path or assertion, the graph can pass unconditionally, required evidence is missing or contradictory, or the relevant path did not run. A visual claim cannot pass without a settled capture. Write only the compact {verdict, reasons} to artifacts/recipe-quality-input.json and build the artifact with mm-harness recipe-quality build; never hand-author it.",
+        "Independent critique, when one is requested: the verdict with one evidence-backed sentence, uncovered or weak proof targets, graph, composition, adapter and reuse findings, evidence mismatches and flake risks, and the three highest-value concrete edits.",
+        {
+          "text": "State the simulator or device, build type and app state. Unlock through a fixture or a documented primitive; keep account, balance, network, permissions and feature flags deterministic; use existing page objects, test ids and fixtures; capture only after ui.wait_for, a route assertion or a state assertion; teardown must not leak balances, pending transactions, selected network, permissions or onboarding state into the next run.",
+          "adapters": [
+            "mobile"
+          ]
+        },
+        {
+          "text": "State browser, build, fixture and the popup, fullscreen or service-worker context. Popup UI proven only through CDP eval or controller reads fails a visual claim; capture after settle; prefer existing e2e and fixture helpers over raw steps.",
+          "adapters": [
+            "extension"
+          ]
+        },
+        {
+          "text": "Core is headless: prove claims with summary.json, trace.json, logs or package tests, never screenshots; use manifest-declared headless actions only; state the target package or workspace, fixture, start state and teardown for controller mutations; a non-visual criterion needs unit or package test coverage.",
+          "adapters": [
+            "core"
+          ]
         }
       ],
       "commands": [