Skip to content

Self-hosted, Self-Improving, Federated Everything App — dev management, agent orchestration, persistent mind, digital twin, second brain, and creative pipeline from a single dashboard over Tailscale to port the user into the OS

fix: make the ux, performance, and mobile-responsive audit prompts fit the runs that execute them

[._.]/ Adam Eivy9d8c06506383
diff --git a/docs/features/self-improvement-audits.md b/docs/features/self-improvement-audits.md
index 847e457d61..570f551c1b 100644
--- a/docs/features/self-improvement-audits.md
+++ b/docs/features/self-improvement-audits.md
@@ -80,7 +80,7 @@ the operator's live checkout.
| `accessibility` | Keyboard, screen-reader, contrast, zoom barriers | `accessibility`, `a11y` | implement | |
| `documentation` | Docs that are wrong, then docs that are missing | `documentation`, `docs` | implement | |
| `ui-bugs` | Things that are broken in the running interface | `bug`, `ui-bug` | implement | |
-| `mobile-responsive` | Small-screen and touch usability | `mobile` | implement | |
+| `mobile-responsive` | Layout at viewport widths ≤1024 and touch mechanics | `mobile` | implement | |
| `error-handling` | Failure paths, timeouts, retries, degraded fallback | `resilience` | implement | |
| `typing` | Type contracts at boundaries | `code-quality`, `typing` | implement | |
| `console-errors` | What the app actually emits while running | `bug`, `console` | implement | |
@@ -118,6 +118,11 @@ problem:
that no metric captures — mixed abstraction levels, flag arguments, names that
lie. `module-hygiene` owns responsibility boundaries and treats a complexity
score only as a candidate signal, never as a finding.
+- **Layout is split by what is narrow.** `mobile-responsive` owns every layout
+ defect at a viewport width ≤1024 (phone and tablet) plus touch mechanics —
+ overflow, clipping, breakpoint gaps, tap targets. `ux` checklist item 8 keeps
+ only a narrow CONTAINER on a desktop viewport (a dashboard widget, an embedded
+ panel, a side panel); the two lanes name each other's side (#12084).
- **Duplication is split two ways.** `simplify` owns copy-paste and dead code.
`better-structural-drift` owns a derived artifact or hand-synced registry acting
as a second source of truth, which is a different fix (derive it, or add the
diff --git a/server/services/taskPromptDefaults.test.js b/server/services/taskPromptDefaults.test.js
index 26d95ef3ea..920723596a 100644
--- a/server/services/taskPromptDefaults.test.js
+++ b/server/services/taskPromptDefaults.test.js
@@ -115,7 +115,7 @@ describe('taskPromptDefaults integrity snapshot', () => {
it('ux filings use tracker IDs instead of title slugs', () => {
const current = DEFAULT_TASK_PROMPTS.ux;
- expect(PROMPT_VERSIONS.ux).toBe(6);
+ expect(PROMPT_VERSIONS.ux).toBe(7);
expect(current).toContain('{trackerInstructions}');
expect(current).toContain('short, human-readable title');
expect(current).not.toContain('[ux-…]');
@@ -145,6 +145,33 @@ describe('taskPromptDefaults integrity snapshot', () => {
expect(current).toContain('**Same kind of page, same frame.**');
});
+ // v7: the harness provisions no Playwright MCP, so the walk names no tool and
+ // no exact viewport; the reflow item keeps only desktop-viewport containers
+ // and hands phone/tablet widths to mobile-responsive (#12077, #12084).
+ it('ux v7 uses whatever browser automation the run has and splits layout by viewport width', () => {
+ const ux = DEFAULT_TASK_PROMPTS.ux;
+ expect(ux).not.toMatch(/Playwright MCP|browser_[a-z]+/);
+ expect(ux).toContain('a CDP tab, Playwright, or the host');
+ expect(ux).toContain('three viewport classes');
+ expect(ux).toContain('(desktop viewport, narrow container)');
+ expect(ux).toContain('Phone and tablet viewports (1024 wide or narrower) and touch mechanics');
+ expect(ux).not.toContain('Responsive layout failures ARE part of this UX audit');
+ expect(PROMPT_VERSIONS['mobile-responsive']).toBe(4);
+ expect(DEFAULT_TASK_PROMPTS['mobile-responsive']).toContain('viewport width ≤1024 and touch');
+ });
+
+ // v6: 'Required' coverage sections made a single-process app pad its report
+ // with 'checked, not applicable' lines (#12078).
+ it('performance v6 scopes platform-specific coverage to what the app has', () => {
+ const perf = DEFAULT_TASK_PROMPTS.performance;
+ expect(PROMPT_VERSIONS.performance).toBe(6);
+ expect(perf).not.toMatch(/^## Required/m);
+ expect(perf.match(/^## Where present:/gm)).toHaveLength(2);
+ expect(perf.match(/skip the rest silently/g)).toHaveLength(2);
+ expect(perf).not.toContain('apps/providers');
+ expect(perf).toContain('**Stable reference data re-downloaded**');
+ });
+
it('ui-bugs v4 hunts unhappy paths and statically provable crashes', () => {
const current = DEFAULT_TASK_PROMPTS['ui-bugs'];
diff --git a/server/services/taskPromptDefaults/integrity.snapshot.json b/server/services/taskPromptDefaults/integrity.snapshot.json
index 7e51ce4b9f..ded1ee229d 100644
--- a/server/services/taskPromptDefaults/integrity.snapshot.json
+++ b/server/services/taskPromptDefaults/integrity.snapshot.json
@@ -35,11 +35,11 @@
"jira-sprint-manager": "3e63c8a3bdf0d5a05a30faaa2dc07980",
"jira-status-report": "9d374a9d8ccb92c5c8fd0aa9899e79a5",
"media-render-research": "5164eda916760e6f3604b0c80a5f33ec",
- "mobile-responsive": "eb7f6b307bccd15199c74d7aff5c21c3",
+ "mobile-responsive": "a1ad3c87dc4c2fc55a3f041f0f61e726",
"model-comparison-refresh": "f66e070fa08e30edda2c3020289f0c67",
"module-hygiene": "e4ba7f0375bf816b85564bcc6545474d",
"observability": "243e58bcfae2759f82caf3ff6c8fc9c5",
- "performance": "8b297d723f9f39154c9cf6d501738494",
+ "performance": "4f8ddada54d7c6f31b7fb35aeb0f29db",
"plan-feature": "7f859397ff16889ffc0cf5a8ba6bddab",
"plan-task": "5dd527c47e736094949c4605cfb2d973",
"plan-task-claim": "7b87285534fc52cc99fa676ecf6939ba",
@@ -61,7 +61,7 @@
"ui-bugs": "06baf650e41fce12aaa59c66030c0357",
"ui-lifecycle": "d156882c73593a54c0fd5754eb4c6fa0",
"user-action-review": "1a083430e43f60cdd586cfffbdebc583",
- "ux": "bb8670c1e6eb28434ad2d45b6dc62177",
+ "ux": "41b02e644f3dc5dcb713700c56a56454",
"workflow-recovery": "628d2ec51c3cd3d8bbd76eacb212dada"
},
"PROMPT_VERSIONS": {
@@ -99,11 +99,11 @@
"jira-sprint-manager": 1,
"jira-status-report": 1,
"media-render-research": 1,
- "mobile-responsive": 3,
+ "mobile-responsive": 4,
"model-comparison-refresh": 4,
"module-hygiene": 2,
"observability": 2,
- "performance": 5,
+ "performance": 6,
"plan-feature": 6,
"plan-task": 21,
"pr-reviewer": 5,
@@ -121,7 +121,7 @@
"ui-bugs": 4,
"ui-lifecycle": 3,
"user-action-review": 2,
- "ux": 6,
+ "ux": 7,
"workflow-recovery": 1
},
"REFERENCE_WATCH_AUDITED_VERSION": 4,
@@ -370,7 +370,8 @@
],
"mobile-responsive": [
"636bdc17fa39ed2f2e27623284c8e04c",
- "16d8e7f63a673a2de48994c0ac08ef3d"
+ "16d8e7f63a673a2de48994c0ac08ef3d",
+ "eb7f6b307bccd15199c74d7aff5c21c3"
],
"model-comparison-refresh": [
"dc1702b07c7fb8be6165feb776d5acb6",
@@ -388,7 +389,8 @@
"4fea5eb3585bdc150980934b289016bc",
"672bd3958b12afb0f382bcebb6dc3b33",
"d7e3a80f42903fbf90be0efc9b559c0b",
- "b69bf4cbd2b3c272b5650608c62cede3"
+ "b69bf4cbd2b3c272b5650608c62cede3",
+ "8b297d723f9f39154c9cf6d501738494"
],
"plan-feature": [
"9fbc2a2ded8f4533d1f0f6b9cb780b34",
@@ -479,7 +481,8 @@
"ac11f692eae40ffe8ddeaeaccbcf75e9",
"b8c0bde4075685625443a405b2c70b50",
"f90b5f38a7c6264e0d40919a2933ba4c",
- "9f346b8be4408c5a8207d9a8b6080ff0"
+ "9f346b8be4408c5a8207d9a8b6080ff0",
+ "bb8670c1e6eb28434ad2d45b6dc62177"
]
}
}
diff --git a/server/services/taskPromptDefaults/prompts.js b/server/services/taskPromptDefaults/prompts.js
index 433289f871..d7bab6c5cf 100644
--- a/server/services/taskPromptDefaults/prompts.js
+++ b/server/services/taskPromptDefaults/prompts.js
@@ -342,11 +342,15 @@ scale the code actually sees. Say which one you have.
import time that only one rare path needs.
- **Cache misuse** — a cache that never invalidates, one that never hits because
its key varies, or an unbounded one that is really a leak.
+- **Stable reference data re-downloaded** — a list that changes only on an
+ event (configuration, providers, a catalog) fetched on a timer or on every
+ visit instead of being invalidated by the event that changes it.
-## Required service and data-path coverage
+## Where present: service and data-path coverage
-For services, APIs, workers, and data pipelines, inspect the heaviest request
-and job paths as well as any UI:
+Apply the items that exist in this app and skip the rest silently. For
+services, APIs, workers, and data pipelines, inspect the heaviest request and
+job paths as well as any UI:
- **Query plans** — read the plan (EXPLAIN or the engine's equivalent) for the
most frequent and the most expensive queries rather than inferring index use
@@ -365,9 +369,10 @@ and job paths as well as any UI:
- **Serialization and startup cost** — repeated encode/decode of the same
payload, and cold-start work on a serverless or per-request path.
-## Required UI load and idle-network coverage
+## Where present: UI load and idle-network coverage
-For apps with a UI, inspect the shared shell and at least one high-volume
+Apply the items that exist in this app and skip the rest silently. For apps
+with a UI, inspect the shared shell and at least one high-volume
collection route, plus a sibling tab that does not use that collection. Trace
initial reads, mounted hidden panels, polling timers, websocket invalidations,
and reconnect recovery. Include the app's largest growing collections (inboxes,
@@ -378,8 +383,7 @@ When browser tooling is available, capture a cold route load, navigation to the
sibling tab, and at least 60 seconds of idle network activity. Report request
counts and transferred/decoded bytes by endpoint, duplicate or overlapping
requests, time until useful content renders, and idle bytes per minute. Include
-websocket payloads, not only fetch/XHR. Check apps/providers and other stable
-reference data are invalidated by events rather than repeatedly downloaded.
+websocket payloads, not only fetch/XHR.
Distinguish expected live telemetry from whole-list polling; telemetry should
be scoped to visible consumers. Use synthetic large collections to verify
server page limits, compact list projections, lazy details, bounded rendered
@@ -803,6 +807,11 @@ tablet width, and a desktop width. Compare: a defect is usually something that
works at one size and fails at another. If the interface is not reachable, exit
cleanly and say so.
+This audit owns every layout defect at a viewport width ≤1024 and touch
+mechanics. A narrow container inside a desktop viewport — a dashboard widget,
+an embedded panel, a side panel — belongs to the UX work's reflow check, so do
+not file it here.
+
## Hunt for
- **Horizontal overflow** — a fixed width, an unwrapped row, a wide table, or a
@@ -858,7 +867,7 @@ interface the same way and hold to the same checklist.
Duplicate findings are noise. Do NOT file:
- **Raw console errors / broken elements / failed requests** — \`ui-bugs\` owns these.
-- **Touch-only mechanics** (hover-only controls and tap-target sizing) — \`mobile-responsive\` owns these. Responsive layout failures ARE part of this UX audit; deduplicate against findings from that sibling task.
+- **Phone and tablet viewports (1024 wide or narrower) and touch mechanics** — \`mobile-responsive\` owns overflow, clipping, breakpoint gaps, viewport-height assumptions, hover-only controls, and tap targets at those viewport widths. This audit's reflow check (item 8) is a narrow CONTAINER on a desktop viewport — a dashboard widget, an embedded panel, a side panel — never a narrow viewport.
- **ARIA labels, contrast ratios, keyboard traps** — \`accessibility\` owns these.
- **The wording of a single string** (jargon, an ambiguous verb, broken plurals) — \`copy\` owns it. How MUCH text a screen spends explaining itself, and whether it repeats what the controls already say, is this audit's (checklist item 9).
- **A workflow that cannot be resumed or cancelled after a failure or reload** — \`workflow-recovery\` owns it. Whether the page tells the user where a record stands is this audit's (checklist item 12).
@@ -888,10 +897,14 @@ keyboard-inaccessible icon") — and file it as the UX finding, not as the a11y
fresh or finished one. Rotate coverage: prefer routes and domains with no
finding from this audit in the tracker in the last 30 days, and list the
routes you chose and why. Describe records by their state, never by their
- names or contents. Walk each chosen route with Playwright MCP. For every route:
- - \`browser_navigate\` to it, then \`browser_snapshot\` to read the structure.
- - \`browser_resize\` to **1440x900**, **1280x800**, **1024x768**, **768x1024**, **390x844**, and **320x812** and
- snapshot at each — the fold differs, and a buried primary action is the
+ names or contents. Walk each chosen route with the browser automation
+ available to this run (a CDP tab, Playwright, or the host's browser tool);
+ if none is reachable, exit cleanly as step 2 says. For every route:
+ - Navigate to it and read the rendered structure (an accessibility snapshot
+ or the DOM), not only a screenshot.
+ - View it in three viewport classes — desktop (1280 wide or more), tablet
+ (roughly 768–1024), and phone (390 or narrower) — and read the structure
+ at each. The fold differs per class, and a buried primary action is the
single most common finding.
- Test with the navigation sidebar expanded and collapsed, and at 200% zoom.
@@ -923,7 +936,7 @@ keyboard-inaccessible icon") — and file it as the UX finding, not as the a11y
7. **Information hierarchy matches the user's task** — what they came for is
the most prominent thing, not the densest table or the newest feature.
- 8. **Narrow-container reflow** — no overlapping cards, clipped controls, off-screen actions, or page-level horizontal scroll. Toolbars wrap, flexible children shrink, long strings wrap or have an accessible full-value path, and side panels stack before squeezing the primary work. Consult the app's design standard; for PortOS use docs/UX_DESIGN_GUIDE.md.
+ 8. **Narrow-container reflow (desktop viewport, narrow container)** — a dashboard widget, embedded panel, or side panel resized to 240–480px wide inside a desktop viewport shows no overlapping cards, clipped controls, off-screen actions, or page-level horizontal scroll. Toolbars wrap, flexible children shrink, long strings wrap or have an accessible full-value path, and side panels stack before squeezing the primary work. The same defects at a phone or tablet viewport width belong to \`mobile-responsive\`. Consult the app's design standard; for PortOS use docs/UX_DESIGN_GUIDE.md.
9. **Text earns its space.** A label, then at most one useful sentence, then
optional detail behind a named disclosure. A finding is a paragraph that
narrates what the visible controls already make obvious ("Click Repair to
diff --git a/server/services/taskPromptDefaults/versions.js b/server/services/taskPromptDefaults/versions.js
index 8fa28eb7b5..bd2e3615a5 100644
--- a/server/services/taskPromptDefaults/versions.js
+++ b/server/services/taskPromptDefaults/versions.js
@@ -38,7 +38,7 @@ export const PROMPT_VERSIONS = {
'security': 6, // v6: generic trust-model pointer (AGENTS.md Security Model + ETHOS.md for PortOS) replaces the PortOS-specific authority paragraph, adds the one-finding-per-class rule, and drops the redaction sentence the mode banners own (#12081, #12080). v5: distinguish delegated authority from untrusted content and require workflow-preserving boundary fixes. v4: adds tenant/resource isolation, SSRF, and unsafe deserialization — the service and data-API defects the v3 list (written for a web app) did not name. v3: carries {modeInstructions} (the pre-v3 body had no placeholder, so its own "fix and commit" steps fought a file-issues dispatch) and defers to the project's documented threat model, requiring an attacker-or-accident story per finding instead of a generic OWASP checklist walk. Previously v2: generic {appName} body (older default hardcoded "PortOS"/"server/routes" paths)
'code-quality': 5, // v5: drops the read-instruction-files and scan-then-bound restatements the mode banners and discovery contract own (#12080). v4: hands generated-artifact/hand-synced-registry/layout-coupling drift to the better-structural-drift lane and counted branching + reader cost to better-complexity / better-cognitive-load, keeping conventional maintainability defects — the split across PortOS's specialist tasks. Previously v3: inventory structural drift, duplicate sources of truth, volatile generated artifacts, and incidental-layout coupling before conventional maintainability checks. v2: generic {appName} body (older default hardcoded "PortOS")
'test-coverage': 3, // v3: carries {modeInstructions}, ranks gaps by consequence rather than percentage, tests at the highest practical public boundary, requires naming the regression a new test uniquely catches, and hands existing-test quality to better-test-quality. Previously v2: generic {appName} body (older default hardcoded "PortOS")
- 'performance': 5, // v5: adds required service/data-path coverage (query plans, streaming vs buffering, batch shape, pools, partition pruning, cold start), hands spend to cost-efficiency, and replaces PortOS-specific collection names with generic ones — the prompt runs against every managed app. v4: requires UI route-load, idle-network and large-collection evidence; missing measurements are explicitly unverified.
+ 'performance': 6, // v6: the service/data-path and UI-load sections are "Where present" (apply what exists, skip the rest silently) and the app-specific reference-data sentence is a generic hunt bullet (#12078). v5: adds required service/data-path coverage (query plans, streaming vs buffering, batch shape, pools, partition pruning, cold start), hands spend to cost-efficiency, and replaces PortOS-specific collection names with generic ones — the prompt runs against every managed app. v4: requires UI route-load, idle-network and large-collection evidence; missing measurements are explicitly unverified.
'accessibility': 5, // v5: the "skip what static analysis already covers" list now also names keyboard-trap/unreachable-control and faked-semantics findings, matching Biome's `useKeyWithClickEvents`, `noNoninteractiveElementToInteractiveRole`, `useFocusableInteractive`, and `noAriaHiddenOnFocusable` a11y rules (enabled in `client/biome.jsonc` alongside the four from v4) — the live-agent budget shifts further to focus-order/live-region/contrast/zoom issues static analysis still can't establish. Previously v4: checks for an existing JSX a11y linter (Biome's a11y group, jsx-a11y) first and skips findings it already catches deterministically (missing alt/labels, invalid/unsupported ARIA), so the live-agent budget goes to keyboard/focus/live-region/contrast/zoom issues static analysis can't establish. Previously v3: carries {modeInstructions}, names who is blocked and how far they get, and prefers a native element over a custom control with patched attributes. Previously v2: generic {appName} + the app UI (older default hardcoded "PortOS" + http://localhost:5555)
'console-errors': 4, // v4: refer to framework-neutral UI lifecycle audit. v3: carries {modeInstructions}, collects real emitted output before diagnosing, groups duplicates by cause, and forbids "fixing" an error by suppressing it. Previously v2: unified `[Improvement: {appName}]` body (pre-unification defaults shipped separate self- and app-improvement bodies)
'error-handling': 4, // v4: cedes sensitive detail in an error path to privacy and partial writes to data-integrity (#12086). v3: carries {modeInstructions} and covers resilience (timeouts, retry ceilings, degraded-dependency fallback, cleanup on the error path, actionable user-facing text) while handing swallowed errors and log context to observability. Previously v2: unified `[Improvement: {appName}]` body (older default used the `[App Improvement: {appName}]` header)
@@ -46,8 +46,8 @@ export const PROMPT_VERSIONS = {
'dependency-updates': 7, // v7: dependency-update worktrees keep package-manager writes out of the source checkout's shared node_modules. v6: the throwaway bot-PR worktree is `dep-{appSlug}-pr-<n>` — the app NAME alone did not isolate two managed apps sharing a name, and {appSlug} is rendered server-side rather than slugified by the agent (`appWorktreeSlug`, server/lib/worktreeOwnership.js). v5: inventory all Dependabot alerts and ecosystems, verify actual upstream fixes, and track unpatched mitigations separately from alert closure. v4: the Phase 1 GitLab MR listing gains `--output json` — without it glab answers with the human table, which carries the head branch but NOT the author, while the classification right below it reads both. The same bump drops `--state` from the MR-list recipes — that flag does not exist (MR state is selected by presence flags `--merged`/`--closed`/`--all`, and OPEN is the default), so `--state opened` exited 1 with `Unknown flag: --state` and Phase 1 skipped its bot-PR triage entirely on GitLab repos. v3: Phase 1 triages open Dependabot/Renovate PRs first (evidence → MERGE / FIX-THEN-MERGE / CLOSE / LEAVE, including conflict + CI-failure repair on the bot branch) before Phase 2 updates anything the bots didn't cover — an agent that bumped packages itself was duplicating and conflicting with open bot PRs. v2: generic {appName} body (older default hardcoded "PortOS")
'documentation': 8, // v8: drops the read-instruction-files restatement the mode banners own (#12080). v7: carries {modeInstructions} and ranks wrong documentation above missing documentation, verifying instructions against the code rather than trusting the prose. Previously v6: the changelog step names `AGENTS.md` (or `CLAUDE.md`) as the file carrying the repo's documented convention — #4852 made AGENTS.md the canonical cross-vendor agent-instructions filename, and a managed app that adopted it would otherwise be told to read a file it no longer has. v5: the changelog step defers to the convention the repo documents (some repos collect per-branch fragments in a directory so parallel agents don't conflict) instead of prescribing an append to `.changelog/NEXT.md`. v4: generic {appName} body (v1 hardcoded "PortOS"; v2/v3 retired DONE.md wording)
'ui-bugs': 4, // v4: hunts the unhappy branch of each primary action and statically provable crashes (a call to a member the imported module lacks), which no longer count as speculative; cedes stranded workflows to workflow-recovery and unfinished controls to feature-completeness. Previous version: v3: carries {modeInstructions}, exits cleanly when the interface is unreachable instead of filing from imagination, and cedes layout/a11y/UX/copy to their owners. Previously v2: generic {appName} + the app UI (older default hardcoded "PortOS" + http://localhost:5555)
- 'mobile-responsive': 3, // v3: carries {modeInstructions} and requires the viewport width plus what a user cannot do there, rather than flagging any layout that merely looks different. Previously v2: generic {appName} app-UI body (older default hardcoded "PortOS" + http://localhost:5555)
- 'ux': 6, // v6: route selection reaches record detail/editor routes partway through their workflow with tracker-driven rotation, and adds checklist items 12 (a record page says where it stands, the next action, and how to open its output) and 13 (same page family, same shell — provable from layout config). Previous version: v5: adds checklist items 9–11 (instruction text that over-explains, duplicate controls for one action, information hidden when the current state needs it) plus a per-route action/text/state inventory; previous runs only filed what items 1–8 named and treated verbosity as aesthetic. v4: forge issues use plain human-readable titles and tracker-defined IDs; filing syntax/identity comes from {trackerInstructions}. v3: audits narrow monitors and widget containers, including clipped overflow and sidebar/zoom states.
+ 'mobile-responsive': 4, // v4: owns every layout defect at viewport width ≤1024 plus touch; narrow containers on a desktop viewport stay with ux (#12084). v3: carries {modeInstructions} and requires the viewport width plus what a user cannot do there, rather than flagging any layout that merely looks different. Previously v2: generic {appName} app-UI body (older default hardcoded "PortOS" + http://localhost:5555)
+ 'ux': 7, // v7: walks routes with whatever browser automation the run has (no Playwright MCP tool names), three viewport classes instead of six exact sizes, and item 8 restricted to a narrow container on a desktop viewport — phone/tablet widths belong to mobile-responsive (#12077, #12084). v6: route selection reaches record detail/editor routes partway through their workflow with tracker-driven rotation, and adds checklist items 12 (a record page says where it stands, the next action, and how to open its output) and 13 (same page family, same shell — provable from layout config). Previous version: v5: adds checklist items 9–11 (instruction text that over-explains, duplicate controls for one action, information hidden when the current state needs it) plus a per-route action/text/state inventory; previous runs only filed what items 1–8 named and treated verbosity as aesthetic. v4: forge issues use plain human-readable titles and tracker-defined IDs; filing syntax/identity comes from {trackerInstructions}. v3: audits narrow monitors and widget containers, including clipped overflow and sidebar/zoom states.
// v2: replaces the hardcoded "Read-only on source / no branches, no PRs" paragraph with {modeInstructions} — the audit is registered as file-issues-capable, so a do-work dispatch previously got a mode banner that its own body contradicted in two places. Previously v1: walk the running UI with Playwright MCP against a 7-item named UX checklist and file ONE tracker item per finding via {trackerInstructions} — read-only on source, no branches/PRs. Mode (file-issues vs implement) is injected at dispatch via {modeInstructions} / applyAuditModeWrapper — no prompt bump required.
'data-safety': 3, // v3: cedes read-modify-write races to data-integrity (#12082). v2: generalizes beyond installs to deployments/environments/datasets and adds backup restorability, bulk/backfill safety, retention reach, and long-lived data formats; hands runtime data correctness to data-integrity. v1: data/upgrade-safety audit (migrations, schema parity, version gates). Mode injected at dispatch.
'simplify': 3, // v3: cedes file splits and responsibility boundaries to module-hygiene (#12083). v2: an unreferenced endpoint/setting/job that documentation, UI text or a plan describes is an unfinished feature for feature-completeness, not a deletion. Previous version: v1: dead-code/duplication audit. Mode injected at dispatch.