omnius 1.0.663 → 1.0.665

package/.aiwg/addons/omnius-docs/skills/browser-interaction-validation/SKILL.md

@@ -7,8 +7,10 @@ compaction_strategy: structured
 # browser-interaction-validation
 
 Validate user-facing browser flows with `playwright_browser` as the canonical
-runtime. A page loading, a successful click receipt, a screenshot, or static
-source assertions alone do **not** prove that a requested interaction worked.
+runtime. A page loading, a successful click receipt, an immediate screenshot,
+or static source assertions alone do **not** prove that a requested interaction
+worked. Some failures appear only after the page remains active. GPU and WebGL
+failures can also differ between headless and headed browsers.
 
 ## Triggers
 
@@ -39,9 +41,12 @@ static checks separately as static evidence.
 1. Use `playwright_browser` for the entire flow. Do not switch to
    `browser_action` (Selenium) mid-flow; it is a separate browser session and
    loses page state. Use Selenium only as an explicit legacy fallback.
-2. Navigate to the live URL, then call `observe_bundle` before interacting.
-   Preserve URL, relevant DOM/a11y selectors, screenshot, page errors, console,
-   and network observations.
+2. Start with `validate_lifecycle`. Give it the live URL, an explicit
+   postcondition, and a finite `observation_ms`. This starts diagnostics before
+   navigation and holds the page open after load. For GUI, GPU, WebGL, WebGPU,
+   canvas, or animation claims, set `headless=false` and `force_new=true`. If a
+   graphical display is unavailable, report the headed check as blocked. Do not
+   substitute a clean headless run as proof.
 3. Choose a stable locator: prefer `data-testid`, then accessible role/name,
    then a specific CSS selector. Avoid ordinal or broad selectors that can race
    reactive re-renders.
@@ -52,9 +57,11 @@ static checks separately as static evidence.
    Re-query after a detached-element/re-render failure; do not retry a stale
    handle blindly. If a stable locator is unavailable, report that as a
    product-testability finding.
-6. Wait for the specific state, then capture post-action `observe_bundle` and
-   a focused DOM/text/a11y observation that proves the postcondition. Include
-   page errors, console errors, and failed network requests in the result.
+6. After the action, call `observe_window` with the expected state and a finite
+   `observation_ms`. Keep the same browser session alive while delayed failures
+   surface. Review the window-scoped page errors, console errors, failed network
+   requests, DOM state, and screenshot. Use `observe_bundle` only for an
+   instantaneous session summary when a timed observation is not required.
 7. Close by listing every requested interaction with its evidence outcome,
    receipt/artifact references, and any precise remediation. Call
    `task_complete` when the implementation work is done; verifier feedback is

package/dist/api/py-embed.js

@@ -289947,7 +289947,6 @@ init_process_lifecycle();
 import { dirname, join, resolve as resolve2 } from "node:path";
 import { homedir as homedir2, tmpdir } from "node:os";
 init_vision();
-var dynamicImport = new Function("mod", "return import(mod)");
 var PLAYWRIGHT_RUNTIME_DIR = join(homedir2(), ".omnius", "playwright-runtime");
 var PLAYWRIGHT_BROWSERS_DIR = join(PLAYWRIGHT_RUNTIME_DIR, "browsers");