diff --git a/.gemini/commands/strict-development-rules.md b/.gemini/commands/strict-development-rules.md
index 3bc00c2717..f1b8364c91 100644
--- a/.gemini/commands/strict-development-rules.md
+++ b/.gemini/commands/strict-development-rules.md
@@ -136,16 +136,12 @@ Gemini CLI project.
 
 ## UI Text & Telemetry
 
-- **Precise Conciseness**: Keep status and error messages under 5 words whenever
-  possible to prevent "information snowblindness."
-- **Telemetry over Etiquette**: Favor status data and raw telemetry over polite
-  filler. Avoid "Please wait," "I'm sorry," or "We're still on it."
-- **Attribution**: Correctly attribute the source of a delay or error:
-    - Use **Gemini** for generative analysis or thinking latency.
-    - Use **System** or **API** for infrastructure, quota, or network issues.
-- **Tone**: Use a friendly, helpful, and humble tone. Use sentence-style
-  capitalization. Solitary sentences aren't punctuated. Speak to the user
-  ("you"). Avoid first-person pronouns ("I" or "we").
+- **UX Writing Standards**: All user-facing strings, error messages, and status
+  telemetry must adhere to the project's UX writing standards.
+- **Reference**: Follow the rules defined in the [string-reviewer skill](../skills/string-reviewer/SKILL.md).
+- **Core Goal**: Prioritize **Precise Conciseness** (under 5 words) and
+  **Telemetry over Etiquette** (raw data over "Please wait").
+
 
 ## Code Cleanup
 
diff --git a/.gemini/skills/string-reviewer/SKILL.md b/.gemini/skills/string-reviewer/SKILL.md
index f37d83b4ad..6bd8eb84be 100644
--- a/.gemini/skills/string-reviewer/SKILL.md
+++ b/.gemini/skills/string-reviewer/SKILL.md
@@ -40,7 +40,9 @@ Use this checklist to audit UI strings and AI responses.
 ### Identity and voice
 - **Eliminate the "I":** Remove all first-person pronouns (I, me, my, mine).
 - **Subject attribution:** Refer to the AI as Gemini and the infrastructure as
-  the - system or the CLI.
+  the system or the CLI. Use visual cues (the Gemini icon/spinner) for 
+  attribution of active thoughts; avoid redundant 3rd-person narrator 
+  prefixes (e.g., "Gemini is thinking about...") in favor of direct telemetry.
 - **Active voice:** Ensure the subject (Gemini or the system) is clearly
   performing the action.
 - **Ownership rule:** Use the system for execution (doing) and Gemini for
diff --git a/docs/get-started/gemini-3.md b/docs/get-started/gemini-3.md
index 8e0af1a9ce..d52161d649 100644
--- a/docs/get-started/gemini-3.md
+++ b/docs/get-started/gemini-3.md
@@ -59,9 +59,8 @@ or fallback to Gemini 2.5 Pro.
 <!-- prettier-ignore -->
 > [!NOTE]
 > The **Keep trying** option uses exponential backoff, in which Gemini
-> CLI waits longer between each retry, when the system is busy. If the retry
-> doesn't happen immediately, please wait a few minutes for the request to
-> process.
+> CLI waits longer between each retry, when the system is busy. Retry after
+> a few minutes if results are not immediate.
 
 ### Model selection and routing types
 
diff --git a/packages/cli/src/ui/components/LoadingIndicator.test.tsx b/packages/cli/src/ui/components/LoadingIndicator.test.tsx
index 089e371c80..9616cba303 100644
--- a/packages/cli/src/ui/components/LoadingIndicator.test.tsx
+++ b/packages/cli/src/ui/components/LoadingIndicator.test.tsx
@@ -50,7 +50,7 @@ const renderWithContext = async (
 
 describe('<LoadingIndicator />', () => {
   const defaultProps = {
-    currentLoadingPhrase: 'Gemini is thinking...',
+    currentLoadingPhrase: 'Thinking...',
     elapsedTime: 5,
   };
 
@@ -71,7 +71,7 @@ describe('<LoadingIndicator />', () => {
     await waitUntilReady();
     const output = lastFrame();
     expect(output).toContain('MockRespondingSpinner');
-    expect(output).toContain('Gemini is thinking...');
+    expect(output).toContain('Thinking...');
     expect(output).toContain('(esc to cancel, 5s)');
   });
 
@@ -108,7 +108,7 @@ describe('<LoadingIndicator />', () => {
 
   it('should display the elapsedTime correctly when Responding', async () => {
     const props = {
-      currentLoadingPhrase: 'Gemini is thinking...',
+      currentLoadingPhrase: 'Thinking...',
       elapsedTime: 60,
     };
     const { lastFrame, unmount, waitUntilReady } = await renderWithContext(
@@ -120,9 +120,9 @@ describe('<LoadingIndicator />', () => {
     unmount();
   });
 
-  it('should display the elapsedTime correctly in human-readable format', async () => {
+  it('should display the elapsedTime correctly in minutes and OMIT seconds for > 1m', async () => {
     const props = {
-      currentLoadingPhrase: 'Gemini is thinking...',
+      currentLoadingPhrase: 'Thinking...',
       elapsedTime: 125,
     };
     const { lastFrame, unmount, waitUntilReady } = await renderWithContext(
@@ -130,7 +130,9 @@ describe('<LoadingIndicator />', () => {
       StreamingState.Responding,
     );
     await waitUntilReady();
-    expect(lastFrame()).toContain('(esc to cancel, 2m 5s)');
+    const output = lastFrame();
+    expect(output).toContain('(esc to cancel, 2m)');
+    expect(output).not.toContain('5s');
     unmount();
   });
 
@@ -229,7 +231,7 @@ describe('<LoadingIndicator />', () => {
   it('should display fallback phrase if thought is empty', async () => {
     const props = {
       thought: null,
-      currentLoadingPhrase: 'Gemini is thinking...',
+      currentLoadingPhrase: 'Thinking...',
       elapsedTime: 5,
     };
     const { lastFrame, unmount, waitUntilReady } = await renderWithContext(
@@ -238,7 +240,7 @@ describe('<LoadingIndicator />', () => {
     );
     await waitUntilReady();
     const output = lastFrame();
-    expect(output).toContain('Gemini is thinking...');
+    expect(output).toContain('Thinking...');
     unmount();
   });
 
@@ -258,15 +260,13 @@ describe('<LoadingIndicator />', () => {
     const output = lastFrame();
     expect(output).toBeDefined();
     if (output) {
-      expect(output).toContain(
-        'Gemini is thinking about Thinking about something...',
-      );
-      expect(output).not.toContain('and other stuff.');
+      expect(output).not.toContain('Gemini is thinking');
+      expect(output).toContain('Thinking about something...');
     }
     unmount();
   });
 
-  it('should use "Gemini is thinking about" if a subject is provided', async () => {
+  it('should NOT prepend "Thinking... " if a subject is provided', async () => {
     const props = {
       thought: {
         subject: 'Planning the response...',
@@ -280,13 +280,12 @@ describe('<LoadingIndicator />', () => {
     );
     await waitUntilReady();
     const output = lastFrame();
-    expect(output).toContain(
-      'Gemini is thinking about Planning the response...',
-    );
+    expect(output).toContain('Planning the response...');
+    expect(output).not.toContain('Thinking... ');
     unmount();
   });
 
-  it('should prioritize thought.subject over currentLoadingPhrase using the new Gemini pattern', async () => {
+  it('should prioritize thought.subject over currentLoadingPhrase', async () => {
     const props = {
       thought: {
         subject: 'This should be displayed',
@@ -301,9 +300,7 @@ describe('<LoadingIndicator />', () => {
     );
     await waitUntilReady();
     const output = lastFrame();
-    expect(output).toContain(
-      'Gemini is thinking about This should be displayed',
-    );
+    expect(output).toContain('This should be displayed');
     expect(output).not.toContain('This should not be displayed');
     unmount();
   });
@@ -351,7 +348,7 @@ describe('<LoadingIndicator />', () => {
       const output = lastFrame();
       // Check for single line output
       expect(output?.trim().includes('\n')).toBe(false);
-      expect(output).toContain('Gemini is thinking...');
+      expect(output).toContain('Thinking...');
       expect(output).toContain('(esc to cancel, 5s)');
       expect(output).toContain('Right');
       unmount();
@@ -375,7 +372,7 @@ describe('<LoadingIndicator />', () => {
       // 3. Right Content
       expect(lines).toHaveLength(3);
       if (lines) {
-        expect(lines[0]).toContain('Gemini is thinking...');
+        expect(lines[0]).toContain('Thinking...');
         expect(lines[0]).not.toContain('(esc to cancel, 5s)');
         expect(lines[1]).toContain('(esc to cancel, 5s)');
         expect(lines[2]).toContain('Right');
@@ -411,7 +408,7 @@ describe('<LoadingIndicator />', () => {
           elapsedTime={5}
           wittyPhrase="I am witty"
           showWit={true}
-          currentLoadingPhrase="Gemini is thinking..."
+          currentLoadingPhrase="Thinking..."
         />,
         StreamingState.Responding,
         120,
@@ -419,7 +416,7 @@ describe('<LoadingIndicator />', () => {
       await waitUntilReady();
       const output = lastFrame();
       // Sequence should be: Primary Text -> Cancel/Timer -> Witty Phrase
-      expect(output).toContain('Gemini is thinking... (esc to cancel, 5s) I am witty');
+      expect(output).toContain('Thinking... (esc to cancel, 5s) I am witty');
       unmount();
     });
 
@@ -429,7 +426,7 @@ describe('<LoadingIndicator />', () => {
           elapsedTime={5}
           wittyPhrase="I am witty"
           showWit={true}
-          currentLoadingPhrase="Gemini is thinking..."
+          currentLoadingPhrase="Thinking..."
         />,
         StreamingState.Responding,
         79,
@@ -443,7 +440,7 @@ describe('<LoadingIndicator />', () => {
       // 3. Witty Phrase
       expect(lines).toHaveLength(3);
       if (lines) {
-        expect(lines[0]).toContain('Gemini is thinking...');
+        expect(lines[0]).toContain('Thinking...');
         expect(lines[1]).toContain('(esc to cancel, 5s)');
         expect(lines[2]).toContain('I am witty');
       }
diff --git a/packages/cli/src/ui/components/LoadingIndicator.tsx b/packages/cli/src/ui/components/LoadingIndicator.tsx
index 673ab8c998..7777af81d9 100644
--- a/packages/cli/src/ui/components/LoadingIndicator.tsx
+++ b/packages/cli/src/ui/components/LoadingIndicator.tsx
@@ -11,7 +11,6 @@ import { theme } from '../semantic-colors.js';
 import { useStreamingContext } from '../contexts/StreamingContext.js';
 import { StreamingState } from '../types.js';
 import { GeminiRespondingSpinner } from './GeminiRespondingSpinner.js';
-import { formatDuration } from '../utils/formatters.js';
 import { useTerminalSize } from '../hooks/useTerminalSize.js';
 import { isNarrowWidth } from '../utils/isNarrowWidth.js';
 import { INTERACTIVE_SHELL_WAITING_PHRASE } from '../hooks/usePhraseCycler.js';
@@ -65,23 +64,27 @@ export const LoadingIndicator: React.FC<LoadingIndicatorProps> = ({
     currentLoadingPhrase === INTERACTIVE_SHELL_WAITING_PHRASE
       ? currentLoadingPhrase
       : thought?.subject
-        ? `Gemini is thinking about ${thoughtLabel ?? thought.subject.trim()}`
+        ? (thoughtLabel ?? thought.subject.trim())
         : currentLoadingPhrase ||
           (streamingState === StreamingState.Responding
-            ? 'Gemini is thinking...'
+            ? 'Thinking...'
             : undefined);
 
   const cancelAndTimerContent =
     showCancelAndTimer &&
     streamingState !== StreamingState.WaitingForConfirmation
-      ? `(esc to cancel, ${elapsedTime < 60 ? `${elapsedTime}s` : formatDuration(elapsedTime * 1000)})`
+      ? `(esc to cancel, ${
+          elapsedTime < 60
+            ? `${elapsedTime}s`
+            : `${Math.floor(elapsedTime / 60)}m`
+        })`
       : null;
 
   const wittyPhraseNode =
     !forceRealStatusOnly &&
     showWit &&
     wittyPhrase &&
-    (primaryText === 'Thinking...' || primaryText === 'Gemini is thinking...') ? (
+    primaryText === 'Thinking...' ? (
       <Box marginLeft={1}>
         <Text color={theme.text.secondary} dimColor italic>
           {wittyPhrase}
diff --git a/packages/cli/src/ui/components/triage/TriageIssues.tsx b/packages/cli/src/ui/components/triage/TriageIssues.tsx
index 62c0f50e1c..b19d158569 100644
--- a/packages/cli/src/ui/components/triage/TriageIssues.tsx
+++ b/packages/cli/src/ui/components/triage/TriageIssues.tsx
@@ -208,8 +208,8 @@ INSTRUCTIONS:
 3. If it seems like a legitimate bug or feature request that needs triage by a human, recommend "keep".
 4. Provide a brief reason for your recommendation.
 5. If recommending "close", provide a polite, professional, and helpful 'suggested_comment' explaining why it's being closed and what the user can do (e.g., provide more logs, follow contributing guidelines).
-6. CRITICAL: If the reason for closing is "Non-deterministic model output", you MUST use the following text EXACTLY as the 'suggested_comment':
-"Thank you for the report. Model outputs are non-deterministic, and we are unable to troubleshoot isolated quality issues that lack a repeatable test case. We are closing this issue while we continue to work on overall model performance and reliability. If you find a way to consistently reproduce this specific issue, please let us know and we can take another look."
+L211- 6. CRITICAL: If the reason for closing is "Non-deterministic model output", you MUST use the following text EXACTLY as the 'suggested_comment':
+"Closing: model outputs are non-deterministic. Re-open with a repeatable test case once available."
 
 Return a JSON object with:
 - "recommendation": "close" or "keep"
diff --git a/packages/core/src/prompts/snippets.ts b/packages/core/src/prompts/snippets.ts
index 27c1fa60a1..66f160cb65 100644
--- a/packages/core/src/prompts/snippets.ts
+++ b/packages/core/src/prompts/snippets.ts
@@ -364,7 +364,8 @@ export function renderOperationalGuidelines(
       ? 'per-tool explanations.'
       : 'mechanical tool-use narration (e.g., "I will now call...").'
   }
-- **Concise & Direct:** Adopt a professional, direct, and concise tone suitable for a CLI environment.
+- **Precise Conciseness:** Prioritize **Precise Conciseness** (under 5 words) and **Telemetry over Etiquette** (raw data over polite filler like "Please wait").
+- **Attribution:** Use visual cues (the Gemini icon/spinner) for attribution of active thoughts; avoid redundant 3rd-person narrator prefixes (e.g., "Gemini is thinking about...") in favor of direct telemetry.
 - **Minimal Output:** Aim for fewer than 3 lines of text output (excluding tool use/code generation) per response whenever practical.
 - **No Chitchat:** Avoid conversational filler, preambles ("Okay, I will now..."), or postambles ("I have finished the changes...") unless they are ${
     options.topicUpdateNarration