- Skills: add a meme-maker skill for curated template search, local SVG/PNG rendering, Imgflip hosted rendering, and Know Your Meme provenance links.

- Skills CLI: allow `openclaw skills install` and `openclaw skills update` to target shared managed skills with `--global`. (#74466) Thanks @Marvae.

- Browser: surface pending and recently handled modal dialogs in snapshots, return `blockedByDialog` when an action opens a modal, and allow `browser dialog --dialog-id` to answer pending dialogs.

- Browser CLI: add `openclaw browser evaluate --timeout-ms` so long-running page functions can extend both the evaluate action and request timeout budgets. (#83447) Thanks @eefreenyc.

- Codex app-server: scope OpenClaw prompt guidance by runtime surface so native Codex keeps Codex-owned base/personality instructions while OpenClaw contributes only runtime context, delivery guidance, and explicitly scoped command hints. (#83454) Thanks @100yenadmin.

- Agents/tools: shorten built-in tool descriptions and schema hints across media, messaging, sessions, cron, Gateway, web, image/PDF, TTS, nodes, and plan tools while preserving routing guardrails.

- Skills: add node inspector debugging, fused diagram generation, and throwaway spike workflow skills.

openclaw browser fill --fields '[{"ref":"1","value":"Ada"}]'

openclaw browser wait --text "Done"

openclaw browser evaluate --fn '(el) => el.textContent' --ref <ref>

openclaw browser evaluate --timeout-ms 30000 --fn 'async () => { await window.ready; return true; }'

```

Use `evaluate --timeout-ms <ms>` when the page-side function may need longer

than the default evaluate timeout.

Action responses return the current raw `targetId` after action-triggered page

replacement when OpenClaw can prove the replacement tab. Scripts should still

store and pass `suggestedTargetId`/labels for long-lived workflows.

openclaw browser wait --text "Done"

openclaw browser wait "#main" --url "**/dash" --load networkidle --fn "window.ready===true"

openclaw browser evaluate --fn '(el) => el.textContent' --ref 7

openclaw browser evaluate --timeout-ms 30000 --fn 'async () => { await window.ready; return true; }'

openclaw browser highlight e12

openclaw browser trace start

openclaw browser trace stop

- `browser act kind=evaluate` / `openclaw browser evaluate` and `wait --fn`

execute arbitrary JavaScript in the page context. Prompt injection can steer

this. Disable it with `browser.evaluateEnabled=false` if you do not need it.

- Use `openclaw browser evaluate --timeout-ms <ms>` when the page-side function

may need longer than the default evaluate timeout.

- For logins and anti-bot notes (X/Twitter, etc.), see [Browser login + X/Twitter posting](/tools/browser-login).

- Keep the Gateway/node host private (loopback or tailnet-only).

- Remote CDP endpoints are powerful; tunnel and protect them.

expect(options?.timeoutMs).toBeGreaterThan(21000);

});

});

describe("browser action input evaluate command", () => {

beforeEach(() => {

mocks.callBrowserRequest.mockClear();

getBrowserCliRuntimeCapture().resetRuntimeCapture();

});

it("passes timeout-ms through to the evaluate action and outer request", async () => {

const program = createActionInputProgram();

await program.parseAsync(

["browser", "evaluate", "--fn", "() => true", "--timeout-ms", "30000"],

{ from: "user" },

);

const request = mocks.callBrowserRequest.mock.calls.at(-1)?.[1] as

| { body?: { timeoutMs?: number } }

| undefined;

const options = mocks.callBrowserRequest.mock.calls.at(-1)?.[2] as

| { timeoutMs?: number }

| undefined;

expect(request?.body?.timeoutMs).toBe(30000);

expect(options?.timeoutMs).toBeGreaterThan(30000);

});

});

.description("Evaluate a function against the page or a ref")

.option("--fn <code>", "Function source, e.g. (el) => el.textContent")

.option("--ref <id>", "Ref from snapshot")

.option(

"--timeout-ms <ms>",

"How long to allow the evaluate function to run (default: 20000)",

(v: string) => Number(v),

)

.option("--target-id <id>", "CDP target id (or unique prefix)")

.action(async (opts, cmd) => {

const { parent, profile } = resolveBrowserActionContext(cmd, parentOpts);

return;

}

try {

const timeoutMs = Number.isFinite(opts.timeoutMs) ? opts.timeoutMs : undefined;

const result = await callBrowserAct<{ result?: unknown }>({

parent,

profile,

fn: opts.fn,

ref: normalizeOptionalString(opts.ref),

targetId: normalizeOptionalString(opts.targetId),

timeoutMs,

},

timeoutMs,

});

if (parent?.json) {

defaultRuntime.writeJson(result);

);

return;

}

const restoredChangedPaths = diffConfigPaths(snapshot.config, restoredMerge.result);

const actor = resolveControlPlaneActor(client);

if (restoredChangedPaths.length === 0) {

context?.logGateway?.info(

`config.patch noop ${formatControlPlaneActor(actor)} (no changed paths)`,

);

respond(

true,

{

ok: true,

noop: true,

path: resolveGatewayConfigPath(snapshot),

config: redactConfigObject(snapshot.config, schemaPatch.uiHints),

},

undefined,

);

return;

}

const validated = validateConfigObjectWithPlugins(restoredMerge.result);

if (!validated.ok) {

respond(

return;

}

const changedPaths = diffConfigPaths(snapshot.config, validated.config);

const actor = resolveControlPlaneActor(client);

// No-op: if the validated config is identical to the current config,

// skip the file write and SIGUSR1 restart entirely. This avoids a full