@rolepod/uiproof 0.4.0 → 0.5.0

package/.claude-plugin/marketplace.json

@@ -1,7 +1,7 @@
 {
   "$schema": "https://anthropic.com/claude-code/marketplace.schema.json",
   "name": "rolepod-uiproof",
-  "description": "Multi-platform UI / mobile automation MCP server + 4 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e) for AI coding agents.",
+  "description": "Multi-platform UI / mobile automation MCP server + 5 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e, check-errors) for AI coding agents.",
   "owner": {
     "name": "nuttaruj",
     "url": "https://github.com/nuttaruj"
@@ -10,8 +10,8 @@
     {
       "name": "rolepod-uiproof",
       "source": "./",
-      "description": "15 MCP tools (10 atomic browser/mobile primitives + 5 composite workflows) + 4 user-invocable skills. Web production-ready via Playwright; mobile (iOS/Android) via Appium scaffolded — see `rolepod-uiproof doctor` for readiness.",
-      "version": "0.4.0",
+      "description": "26 MCP tools (21 atomic browser/mobile primitives + 5 composite workflows) + 5 user-invocable skills. v0.5 completes the UI verification surface — console + network observability, hover/drag/fill_form/upload/dialog, runtime emulation, multi-page, gated JS eval — replacing chrome-devtools-mcp and playwright-mcp for UI testing. Web production-ready via Playwright; mobile (iOS/Android) via Appium scaffolded — see `rolepod-uiproof doctor` for readiness.",
+      "version": "0.5.0",
       "author": {
         "name": "nuttaruj"
       },

package/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "rolepod-uiproof",
-  "version": "0.4.0",
-  "description": "Multi-platform UI/mobile automation for AI agents — 4 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e) + MCP server with 15 tools. v0.3 adds AppiumEngine scaffolding for iOS/Android, scope={ref} audit, replay CLI, ddmin minimization, doctor + install:mobile.",
+  "version": "0.5.0",
+  "description": "Multi-platform UI/mobile automation for AI agents — 5 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e, check-errors) + MCP server with 26 tools. v0.5 completes the UI verification surface: console + network observability, hover/drag/fill_form/upload/dialog, runtime emulation, multi-page support, gated JS eval — replacing chrome-devtools-mcp and playwright-mcp for UI testing.",
   "author": {
     "name": "nuttaruj",
     "url": "https://github.com/nuttaruj"
@@ -19,7 +19,7 @@
   "mcpServers": {
     "rolepod-uiproof": {
       "command": "npx",
-      "args": ["-y", "@rolepod/uiproof"],
+      "args": ["-y", "rolepod-uiproof"],
       "env": {}
     }
   }

package/.codex-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "rolepod-uiproof",
-  "version": "0.4.0",
-  "description": "Multi-platform UI/mobile automation for AI agents — 4 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e) + MCP server with 15 tools. v0.3 adds AppiumEngine scaffolding for iOS/Android; web is production-ready via Playwright.",
+  "version": "0.5.0",
+  "description": "Multi-platform UI/mobile automation for AI agents — 5 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e, check-errors) + MCP server with 26 tools. v0.5 completes the UI verification surface: console + network observability, hover/drag/fill_form/upload/dialog, runtime emulation, multi-page, gated JS eval.",
   "author": {
     "name": "nuttaruj",
     "url": "https://github.com/nuttaruj"
@@ -21,17 +21,11 @@
     "visual-regression"
   ],
   "skills": "./skills/",
-  "mcp_servers": {
-    "rolepod-uiproof": {
-      "command": "npx",
-      "args": ["-y", "@rolepod/uiproof"],
-      "env": {}
-    }
-  },
+  "mcpServers": "./.mcp.json",
   "interface": {
     "displayName": "Rolepod UIProof",
     "shortDescription": "UI verification, a11y audits, visual diff, e2e scaffolding — for AI coding agents.",
-    "longDescription": "rolepod-uiproof ships an MCP server with 15 tools (10 atomic + 5 composite) and 4 user-invocable skills (/verify-ui, /audit-a11y, /visual-diff, /scaffold-e2e). Web is fully supported via Playwright; mobile (iOS/Android via Appium) lands in v0.3.",
+    "longDescription": "rolepod-uiproof ships an MCP server with 26 tools (21 atomic + 5 composite) and 5 user-invocable skills (/verify-ui, /audit-a11y, /visual-diff, /scaffold-e2e, /check-errors). Web is fully supported via Playwright; mobile (iOS/Android via Appium) supports basic input. v0.5 completes the UI verification surface — console logs, network requests, HAR/video/trace capture, runtime emulation, multi-page popups.",
     "developerName": "nuttaruj",
     "category": "Productivity",
     "capabilities": ["Read", "Write", "Bash"],

package/.cursor/mcp.json

@@ -2,7 +2,7 @@
   "mcpServers": {
     "rolepod-uiproof": {
       "command": "npx",
-      "args": ["-y", "@rolepod/uiproof"],
+      "args": ["-y", "rolepod-uiproof"],
       "env": {}
     }
   }

package/.cursor-plugin/plugin.json

@@ -1,8 +1,8 @@
 {
   "name": "rolepod-uiproof",
   "displayName": "Rolepod UIProof",
-  "version": "0.4.0",
-  "description": "Multi-platform UI / mobile automation MCP server + 4 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e) for AI coding agents. Web production-ready via Playwright; mobile (iOS/Android via Appium) scaffolded.",
+  "version": "0.5.0",
+  "description": "Multi-platform UI / mobile automation MCP server + 5 shipped skills (verify-ui, audit-a11y, visual-diff, scaffold-e2e, check-errors) for AI coding agents. v0.5 completes the UI verification surface: console + network observability, hover/drag/fill_form/upload/dialog, runtime emulation, multi-page, gated JS eval — replacing chrome-devtools-mcp and playwright-mcp.",
   "author": {
     "name": "nuttaruj"
   },

package/CHANGELOG.md

@@ -7,6 +7,127 @@ release.
 
 ## [Unreleased]
 
+## [0.5.0] — 2026-05-27
+
+**Complete UI verification surface — one MCP replaces chrome-devtools-mcp
+and playwright-mcp for UI testing use cases.**
+
+Tool count: 15 → 26 (atomic 10 → 21, composite 5 unchanged). The five
+"out of scope for `uiproof`" families (Lighthouse, performance traces,
+heap snapshots, extensions, third-party page tools) are intentionally
+**not** added — those belong to future `rolepod-perfproof` and
+`rolepod-secproof` MCPs.
+
+### Added — 11 new atomic tools
+
+Cross-platform (work on chromium/firefox/webkit; mobile stubs throw
+`engine_error` until gestures land):
+
+- `rolepod_browser_hover` — `locator.hover()`; refs stay valid
+- `rolepod_browser_drag` — `locator.dragTo()`
+- `rolepod_browser_fill_form` — batch input/select/checkbox/radio
+- `rolepod_browser_upload_file` — `locator.setInputFiles()`, abs path required
+
+Web-only (cast to `PlaywrightEngine`):
+
+- `rolepod_browser_handle_dialog` — pre-arm one-shot accept/dismiss
+- `rolepod_browser_console` — list/filter/clear ring-buffered console
+  messages (1000-entry cap, errors+warnings default)
+- `rolepod_browser_network` — list/filter network requests, optional HAR export
+- `rolepod_browser_set_env` — runtime viewport / offline / geolocation /
+  color_scheme / reduced_motion / extra_headers / network_throttle (CDP) /
+  cpu_throttle (CDP)
+- `rolepod_browser_evaluate` — arbitrary JS in page context.
+  **Disabled by default** — opt in via `ROLEPOD_ALLOW_EVAL=1` env var
+- `rolepod_browser_pages` — list pages in active context (popups,
+  target=_blank, OAuth windows)
+- `rolepod_browser_switch_page` — set active page index
+
+### Added — verify_ui_flow capture lifecycle (impl)
+
+The `capture` array has accepted these values since v0.1, but only
+`screenshot` was wired. v0.5 fills in the rest:
+
+- `console` → `{runDir}/console.json`
+- `har` → `{runDir}/network.har`
+- `video` → `{runDir}/videos/*.webm`
+- `trace` → `{runDir}/trace.zip` (view with `npx playwright show-trace`)
+- `a11y_tree` → `{runDir}/a11y_tree.json`
+
+### Added — 8 new verify_ui_flow step kinds
+
+`hover`, `drag`, `fill_form`, `upload`, `dialog`, `set_env`,
+`switch_page`, `evaluate`. All get first-class codegen in
+`scaffold_e2e` for playwright-test and pytest+selenium.
+
+### Added — 4 new verify_ui_flow expect kinds
+
+- `no_console_errors` — filter level=error, drop excludes, count must be 0
+- `no_failed_requests` — filter `failure || status>=400` (or `>=500`
+  when `allow_4xx`), drop excludes, count must be 0
+- `request_made` — URL regex + optional method must match `min_count`
+  (default 1) times
+- `response_status` — URL regex + exact status code must match
+
+### Added — multi-page support
+
+A session is now a `context` (was a single page). Popups and
+`target="_blank"` links are auto-tracked. Use `browser_pages` to list,
+`browser_switch_page` to activate. Default active = page 0.
+
+### Added — new skill `/check-errors`
+
+Thin wrapper over `rolepod_verify_ui_flow` with strict assertions baked
+in. Use case: PR-gate or post-merge smoke.
+
+### Changed — `/verify-ui` and `/scaffold-e2e` skills
+
+Documented every new step / expect / capture kind. Default suggestion
+in `/verify-ui`: include `no_console_errors` and `no_failed_requests`
+in `expect` for any user-visible flow.
+
+### Changed — Engine interface
+
+Adds four cross-platform input methods: `hover`, `drag`, `fillForm`,
+`uploadFile`. `OpenOptions.capture` accepts `{ har, video, trace }`.
+`WebSession.page` renamed to `mainPage`; internal call sites go through
+`activePage(s)`.
+
+### Non-changes (intentional)
+
+- `screencast_*` not added — Playwright `trace.zip` is strictly better.
+- `click_at` not added — use refs from `snapshot`.
+- Lighthouse not added — axe-core covers a11y.
+- Performance traces / heap snapshots not added — `rolepod-perfproof` scope.
+- Extension management not added — out of scope.
+
+### Migration from 0.4
+
+Pure additions; no behavioral changes on existing tools or
+step/expect/capture kinds. Existing replay bundles play back unchanged.
+
+## [0.4.1] — 2026-05-27
+
+### Fixed
+
+- **MCP server failed to start via `npx -y @rolepod/uiproof`** — npx
+  resolves a scoped package's default executable by the *short* name
+  (`uiproof`, after the `/`), but our `bin` field only defined
+  `rolepod-uiproof`. The mismatch caused `sh: rolepod-uiproof: command
+  not found` and an MCP `-32000` reconnect error in Claude Code,
+  Cursor, etc. All `mcpServers` entries (`.mcp.json`, `.cursor/mcp.json`,
+  `.claude-plugin/plugin.json`, `plugins/rolepod-uiproof/.mcp.json`)
+  now invoke `npx -y rolepod-uiproof`, which npx resolves directly by
+  bin name from the registry.
+
+### Added
+
+- Tool `title` and `annotations` (`readOnlyHint` / `destructiveHint` /
+  `idempotentHint` / `openWorldHint`) on all 15 tools, per the MCP
+  2025-11-25 spec. Clients use these hints to render friendlier names
+  in the picker and to auto-approve read-only calls vs. prompting
+  harder on destructive ones.
+
 ## [0.4.0] — 2026-05-24
 
 **Breaking: project renamed `rolepod-mcp` → `rolepod-uiproof`.**

package/README.md

@@ -1,25 +1,29 @@
 # rolepod-uiproof
 
-**rolepod-uiproof gives Claude Code, Cursor, Codex CLI, and Gemini CLI a real browser/mobile driver — so the AI can actually click through your UI, audit accessibility, diff screenshots, and scaffold e2e tests instead of guessing.**
+**rolepod-uiproof gives Claude Code, Cursor, Codex CLI, and Gemini CLI a real browser/mobile driver — so the AI can actually click through your UI, audit accessibility, check console errors, inspect network requests, diff screenshots, and scaffold e2e tests instead of guessing.**
 
-One MCP server, one tool surface, four skills you invoke from chat. Web is production-ready via Playwright; iOS and Android use Appium (same client as alumnium — needs a local Appium daemon + simulator/emulator, or a real device). No internal LLM — your Lead agent drives every action.
+One MCP server, one tool surface, five skills you invoke from chat. Web is production-ready via Playwright; iOS and Android use Appium (same client as alumnium — needs a local Appium daemon + simulator/emulator, or a real device). No internal LLM — your Lead agent drives every action.
+
+**v0.5 completes the UI verification surface — replacing `chrome-devtools-mcp` and `playwright-mcp` for UI testing.** 26 tools total (21 atomic + 5 composite). New in v0.5: console + network observability, hover / drag / fill_form / upload / dialog, runtime emulation (resize / offline / geolocation / color_scheme / network + CPU throttle), multi-page support, gated JS eval, and impl of HAR / video / trace capture in `/verify-ui`.
 
 ## What it helps with
 
-- **Verify a UI change in seconds.** `/verify-ui` opens a real browser, runs your steps, checks your assertions, saves a screenshot + replay bundle.
+- **Verify a UI change in seconds.** `/verify-ui` opens a real browser, runs your steps, checks your assertions, saves a screenshot + replay bundle (optionally HAR + video + trace + console logs).
+- **Gate merges on "no regressions during this flow".** `/check-errors` runs a flow with strict `no_console_errors` + `no_failed_requests` assertions baked in. PR-gate or post-merge smoke check.
 - **Catch a11y regressions before merge.** `/audit-a11y` runs axe-core against WCAG-A / AA / AAA and returns issues grouped by severity, with WCAG references and fix links.
 - **Lock down the visual contract.** `/visual-diff` captures a screenshot and compares against a named baseline under `./.rolepod-uiproof/baselines/`. First call seeds; subsequent calls diff.
-- **Turn an interactive verify run into a real test file.** `/scaffold-e2e` transcribes a replay bundle into Playwright Test, Vitest+Playwright, or pytest+selenium.
+- **Turn an interactive verify run into a real test file.** `/scaffold-e2e` transcribes a replay bundle into Playwright Test, Vitest+Playwright, or pytest+selenium — with first-class codegen for every step + expect kind.
 - **Reproduce + minimize a bug deterministically.** `/verify-ui` with `mode: "reproduce"` runs ddmin step-elimination to find the shortest still-reproducing sequence.
 
-## The four skills
+## The five skills
 
 | Skill | Wraps | What it does |
 |---|---|---|
-| `/verify-ui` | `rolepod_verify_ui_flow` | Drive a session through steps, evaluate assertions, save evidence + replay bundle. `mode: assert` (default) or `reproduce` with optional ddmin minimization. |
+| `/verify-ui` | `rolepod_verify_ui_flow` | Drive a session through steps, evaluate assertions (incl. console errors / failed requests / specific request made / response status), save evidence (screenshot / console / HAR / video / trace / a11y_tree) + replay bundle. `mode: assert` or `reproduce` with optional ddmin minimization. |
+| `/check-errors` | `rolepod_verify_ui_flow` | Thin wrapper with strict `no_console_errors` + `no_failed_requests` baked in. Use as PR-gate or post-merge smoke. |
 | `/audit-a11y` | `rolepod_audit_a11y` | axe-core audit at WCAG-A / AA / AAA. `scope: "page"` or `scope: { ref }`. Markdown or JSON report. |
 | `/visual-diff` | `rolepod_visual_diff` | Pixel diff against a named baseline. Auto-seeds on first call. Configurable threshold + pixelmatch sensitivity. |
-| `/scaffold-e2e` | `rolepod_scaffold_e2e` | Generate a runnable test file from a scenario + optional replay bundle. Three target frameworks. |
+| `/scaffold-e2e` | `rolepod_scaffold_e2e` | Generate a runnable test file from a scenario + optional replay bundle. Three target frameworks. v0.5 codegen handles every step + expect kind. |
 
 Every skill is **single-backend** (D-024) — it calls the rolepod-uiproof server and only the rolepod-uiproof server. If the server is unavailable, the skill fails with a clear diagnostic. Multi-backend routing belongs in the parent [`rolepod`](https://github.com/nuttaruj/rolepod) plugin's phase skills, not here.
 
@@ -92,7 +96,7 @@ Use this when your tool reads a standard `mcpServers` config (most non-CLI MCP c
   "mcpServers": {
     "rolepod-uiproof": {
       "command": "npx",
-      "args": ["-y", "@rolepod/uiproof"]
+      "args": ["-y", "rolepod-uiproof"]
     }
   }
 }