cc-viewer 1.6.167 → 1.6.169

package/README.md

@@ -125,6 +125,8 @@ You can even scan a QR code to code from your mobile device:
 
 Fulfill your imagination of mobile programming. There's also a plugin mechanism — if you need to customize for your coding habits, stay tuned for plugin hooks updates.
 
+**Voice input:** tap the microphone icon in the chat input for voice-to-text (Web Speech API; requires HTTPS or localhost, so it's hidden on LAN HTTP access). Alternatively, on Android use Gboard's built-in 🎤 key, and on iOS use the system dictation on the keyboard — both work offline and without any HTTPS requirement.
+
 ### Logger Mode (View Complete Claude Code Sessions)
 
 <img width="1500" height="768" alt="image" src="https://github.com/user-attachments/assets/a8a9f3f7-d876-4f6b-a64d-f323a05c4d21" />

package/cli.js

@@ -6,7 +6,7 @@ import { fileURLToPath } from 'node:url';
 import { homedir } from 'node:os';
 import { spawn } from 'node:child_process';
 import { t } from './i18n.js';
-import { INJECT_IMPORT, resolveCliPath, resolveNativePath, resolveNpmClaudePath, buildShellCandidates, setLogDir, LOG_DIR, hasClaude2xWrapper, getGlobalNodeModulesDir, PACKAGES } from './findcc.js';
+import { INJECT_IMPORT, resolveCliPath, resolveNativePath, resolveNpmClaudePath, buildShellCandidates, setLogDir, LOG_DIR, hasClaude2xWrapper, getGlobalNodeModulesDir, PACKAGES, getClaudeConfigDir } from './findcc.js';
 import { ensureHooks } from './lib/ensure-hooks.js';
 
 const __dirname = fileURLToPath(new URL('.', import.meta.url));
@@ -274,6 +274,10 @@ async function runProxyCommand(args) {
       }
     });
 
+    // 注入默认 --thinking-display summarized（用户已显式传入则不覆盖）
+    const { withDefaultThinkingDisplay } = await import('./pty-manager.js');
+    cmdArgs = withDefaultThinkingDisplay(cmdArgs);
+
     cmdArgs.unshift(settingsJson);
     cmdArgs.unshift('--settings');
 
@@ -650,7 +654,7 @@ if (isUninstall) {
 
   // 清理 statusLine 配置和脚本（兼容历史版本遗留）
   try {
-    const settingsPath = resolve(homedir(), '.claude', 'settings.json');
+    const settingsPath = resolve(getClaudeConfigDir(), 'settings.json');
     if (existsSync(settingsPath)) {
       const settings = JSON.parse(readFileSync(settingsPath, 'utf-8'));
       if (settings.statusLine?.command?.includes('ccv-statusline')) {
@@ -659,13 +663,13 @@ if (isUninstall) {
         console.log('Cleaned statusLine config from settings.json');
       }
     }
-    const ccvScript = resolve(homedir(), '.claude', 'ccv-statusline.sh');
+    const ccvScript = resolve(getClaudeConfigDir(), 'ccv-statusline.sh');
     if (existsSync(ccvScript)) {
       unlinkSync(ccvScript);
       console.log('Removed ccv-statusline.sh');
     }
     // 清理 context-window.json
-    const ctxFile = resolve(homedir(), '.claude', 'context-window.json');
+    const ctxFile = resolve(getClaudeConfigDir(), 'context-window.json');
     if (existsSync(ctxFile)) {
       unlinkSync(ctxFile);
     }

package/concepts/en/CustomUltraplanExpert.md

@@ -0,0 +1,161 @@
+# Custom UltraPlan Expert — Authoring Guide
+
+## What the two input fields do
+
+- **Expert name**: the label shown on the role button in the UltraPlan variant row (30-char max). It's just a display name and is **never** sent to Claude Code.
+- **Prompt body**: your role instruction. At send time cc-viewer **automatically** wraps it in `<system-reminder>...</system-reminder>` tags with a `[SCOPED INSTRUCTION]` scope header. So **write the body only** — do not add `<system-reminder>` tags yourself.
+
+---
+
+## What does the expert template look like?
+
+Every built-in expert (Code Expert / Research Expert) is essentially a `<system-reminder>` block injected into Claude Code's context. Your custom expert goes through the exact same pipeline. Here's the **Research Expert** template broken down:
+
+```xml
+<system-reminder>
+[SCOPED INSTRUCTION] The following instructions are intended for the next 1–3
+interactions. Once the task is complete, these instructions should be gradually
+deprioritized and no longer influence subsequent interactions.
+
+Leverage a multi-agent exploration mechanism to formulate an exceptionally
+detailed implementation plan.
+
+Instructions:
+1. Utilize the Agent tool to spawn parallel agents that simultaneously explore
+   various facets of the requirements:
+   - If necessary, deploy a preliminary investigator to conduct an initial
+     survey of industry-specific solutions using `webSearch`;
+   - If necessary, deploy a specialized investigator to research authoritative
+     sources—such as academic papers, news articles, and research reports—
+     using `webSearch`;
+   - Assign an agent to synthesize the target solution, while simultaneously
+     verifying the rigor and credibility of the gathered papers, news, and
+     research reports;
+   - If necessary, assign an agent to analyze competitor data to provide
+     supplementary analytical perspectives;
+   - If necessary, assign an agent to handle the implementation of a product
+     demo (generating outputs such as HTML, Markdown, etc.);
+   - If the task is sufficiently complex, you may assign additional teammates
+     to the roles defined above, or introduce other specialized roles; you are
+     permitted to schedule up to 5 teammates concurrently.
+
+2. Synthesize the findings from the aforementioned agents into a comprehensive,
+   step-by-step implementation plan.
+
+3. Utilize the Agent tool to spawn a set of parallel review agents; these
+   agents shall scrutinize the plan from multiple roles and perspectives to
+   identify any omitted steps and to propose reasonable additions or
+   optimizations.
+
+4. Consolidate the feedback received from the review agents, then invoke
+   `ExitPlanMode` to submit your final plan.
+
+5. Upon receiving the result from `ExitPlanMode`:
+   - If Approved: Proceed to execute the plan within this current session.
+   - If Rejected: Revise the plan based on the provided feedback, and then
+     invoke `ExitPlanMode` once again.
+   - If an Error Occurs: Do *not* follow the suggestions; prompt the user for
+     further instructions.
+
+Your final plan must include the following elements:
+- A clear summary of the proposed implementation strategy;
+- An ordered list of files to be created or modified, specifying the exact
+  changes required for each;
+- A step-by-step sequence for executing the implementation;
+- Identification of potential risks and corresponding mitigation strategies;
+- Creative ideation and suggestions for advanced enhancements;
+- If a product demo was generated, place the corresponding demo output in an
+  appropriate location and notify the user.
+</system-reminder>
+```
+
+---
+
+## Section-by-section breakdown
+
+### 1. `[SCOPED INSTRUCTION]` scope header (wrapper — auto-generated)
+> The following instructions are intended for the next 1–3 interactions...
+
+This tells Claude Code: **these instructions are only active for the next 1–3 turns**, then fade out. Prevents the "expert persona" from leaking into unrelated conversation afterwards.
+
+**This line is generated by cc-viewer automatically. You don't need to write it.**
+
+### 2. Opening task definition (**this is what you should rewrite**)
+> Leverage a multi-agent exploration mechanism to formulate an exceptionally detailed implementation plan.
+
+This is the "subject-verb-object" of the whole template: **it tells Claude Code the posture and objective**. The default "multi-agent exploration + implementation plan" fits **software engineering / planning** tasks well, but feels awkward for many other domains (content review, data analysis, copywriting, market research, compliance audit…).
+
+**We strongly recommend rewriting this line for your objective**, for example:
+
+- **Content reviewer**: "You are a senior content reviewer specializing in {domain}. Your goal is to identify factual inaccuracies, tone inconsistencies, and structural weaknesses in the provided material."
+- **Competitive analyst**: "Conduct a rigorous competitive analysis for {product category}. Produce a comparison matrix, positioning insights, and strategic recommendations."
+- **Copywriter**: "Generate multiple creative copy variants for {scenario}, each with distinct positioning, tone, and call-to-action strategy."
+
+### 3. Workflow steps (1–5 items — **trim or extend based on complexity**)
+
+The Research Expert has 5 steps: **explore → synthesize → review → submit plan → execute**. This enforces "parallel multi-agent + cross-review + plan approval" — three layers of rigor, appropriate for high-stakes/wide-scope tasks but **overkill for lightweight ones**.
+
+- **Simple task** (single lookup / small fix): drop the multi-agent dispatch and review; just "produce the answer" in one step.
+- **Moderate task**: keep "explore → synthesize → review"; drop the ExitPlanMode dance; deliver the result directly.
+- **Complex, high-cost task** (large refactor, multi-option comparison, cross-domain research): keep all 5 steps, possibly add a "risk model" or "option-comparison matrix" step.
+
+### 4. Sub-roles in Step 1 (**tailor to your domain**)
+
+Research Expert lists 6 potential roles (industry scout, academic researcher, synthesizer + fact-checker, competitor analyst, demo producer, extensibility slot). **Rewrite this list for your scenario**:
+
+- **Writing**: "source collector + style analyst + fact checker"
+- **Data analysis**: "data-cleaning agent + statistical modeling agent + visualization agent"
+- **Code audit**: "static-analysis agent + dependency-chain auditor + threat modeler"
+
+### 5. Final deliverable checklist (**align with your real need**)
+
+> Your final plan must include the following elements: ...
+
+The original template lists 6 elements of an "implementation plan". Your deliverable could be something else entirely:
+
+- A **research report** → "Executive summary / Methodology / Key findings / Limitations / Action recommendations"
+- A **review report** → "Issue list / Severity rating / Fix suggestions / Before-after examples"
+- A **comparison matrix** → "Dimension definitions / Scoring rubric / Conclusions / Recommendation rationale"
+
+---
+
+## Authoring tips (TL;DR)
+
+1. **Keep the wrapper**: `<system-reminder>` + `[SCOPED INSTRUCTION]` line is added by cc-viewer — don't repeat.
+2. **Rewrite the opening sentence**: state the role, goal, and output format in one line.
+3. **Flex the workflow**: 1–2 steps for light tasks, the full 5-step loop only for complex ones.
+4. **Rewrite Step 1 sub-roles**: the defaults (academic papers / competitors / demo) are probably not what you want.
+5. **The final "deliverable checklist" is your quality bar**: spell out the output structure — Claude Code will follow it strictly.
+
+---
+
+## A refactored example: Competitive Analyst
+
+```
+You are a senior competitive intelligence analyst for {industry}. Your goal is to
+produce a decision-grade competitive landscape report for the product "{our product}".
+
+Instructions:
+1. Use the Agent tool to dispatch 3 parallel investigators:
+   - Market landscape agent: map the top 5–8 competitors with core positioning
+   - Feature matrix agent: compile a feature-by-feature comparison using
+     publicly available sources (webSearch)
+   - Pricing & GTM agent: analyze pricing models, distribution channels, and
+     go-to-market motions
+
+2. Synthesize the three streams into a unified competitive report.
+
+3. Dispatch one review agent to stress-test the report: challenge any
+   assumption lacking a cited source, flag outdated data (>12 months), and
+   propose one "non-obvious" insight.
+
+4. Deliver the final report with the following sections:
+   - TL;DR (3 bullets)
+   - Competitor positioning map
+   - Feature matrix (markdown table)
+   - Pricing & GTM table
+   - Top 3 strategic implications for our product
+   - Caveats & data gaps
+```
+
+Compared to the original Research Expert: trimmed to 4 steps, sub-roles reduced from 6 to 3, deliverable list fully rewritten as "report sections".

package/concepts/zh/CustomUltraplanExpert.md

@@ -0,0 +1,158 @@
+# 自定义 UltraPlan 专家 — 使用说明
+
+## 两个输入框的作用
+
+- **专家名称**：显示在 UltraPlan 弹窗变体行里的按钮标题，30 字符以内。只是一个标签，不会被发送给 Claude Code。
+- **提示词正文**：你写给 Claude Code 的"角色指令"。发送时会被**自动**用 `<system-reminder>...</system-reminder>` 标签包裹，并在最前面加上一行 `[SCOPED INSTRUCTION]` 的作用域声明。所以你**只需要写正文**，不要自己加 `<system-reminder>` 标签。
+
+---
+
+## 专家模版是什么样的
+
+所有内置专家（代码专家 / 调研专家）本质上都是一段注入到 Claude Code 上下文里的 `<system-reminder>`。你的自定义专家会走一模一样的管道。下面以**调研专家**为例拆解给你看：
+
+```xml
+<system-reminder>
+[SCOPED INSTRUCTION] The following instructions are intended for the next 1–3
+interactions. Once the task is complete, these instructions should be gradually
+deprioritized and no longer influence subsequent interactions.
+
+Leverage a multi-agent exploration mechanism to formulate an exceptionally
+detailed implementation plan.
+
+Instructions:
+1. Utilize the Agent tool to spawn parallel agents that simultaneously explore
+   various facets of the requirements:
+   - If necessary, deploy a preliminary investigator to conduct an initial
+     survey of industry-specific solutions using `webSearch`;
+   - If necessary, deploy a specialized investigator to research authoritative
+     sources—such as academic papers, news articles, and research reports—
+     using `webSearch`;
+   - Assign an agent to synthesize the target solution, while simultaneously
+     verifying the rigor and credibility of the gathered papers, news, and
+     research reports;
+   - If necessary, assign an agent to analyze competitor data to provide
+     supplementary analytical perspectives;
+   - If necessary, assign an agent to handle the implementation of a product
+     demo (generating outputs such as HTML, Markdown, etc.);
+   - If the task is sufficiently complex, you may assign additional teammates
+     to the roles defined above, or introduce other specialized roles; you are
+     permitted to schedule up to 5 teammates concurrently.
+
+2. Synthesize the findings from the aforementioned agents into a comprehensive,
+   step-by-step implementation plan.
+
+3. Utilize the Agent tool to spawn a set of parallel review agents; these
+   agents shall scrutinize the plan from multiple roles and perspectives to
+   identify any omitted steps and to propose reasonable additions or
+   optimizations.
+
+4. Consolidate the feedback received from the review agents, then invoke
+   `ExitPlanMode` to submit your final plan.
+
+5. Upon receiving the result from `ExitPlanMode`:
+   - If Approved: Proceed to execute the plan within this current session.
+   - If Rejected: Revise the plan based on the provided feedback, and then
+     invoke `ExitPlanMode` once again.
+   - If an Error Occurs: Do *not* follow the suggestions; prompt the user for
+     further instructions.
+
+Your final plan must include the following elements:
+- A clear summary of the proposed implementation strategy;
+- An ordered list of files to be created or modified, specifying the exact
+  changes required for each;
+- A step-by-step sequence for executing the implementation;
+- Identification of potential risks and corresponding mitigation strategies;
+- Creative ideation and suggestions for advanced enhancements;
+- If a product demo was generated, place the corresponding demo output in an
+  appropriate location and notify the user.
+</system-reminder>
+```
+
+---
+
+## 逐段解释
+
+### 1. `[SCOPED INSTRUCTION]` 作用域声明（外壳，自动生成）
+> The following instructions are intended for the next 1–3 interactions...
+
+这行告诉 Claude Code：这段指令**只在未来 1–3 轮对话中强制执行**，任务完成后就淡出，不再影响后续交互。避免"专家人格"污染到任务之外的对话。
+
+**这行由 ccv 自动生成，你不需要写。**
+
+### 2. 首段任务定义（**这段最值得你改**）
+> Leverage a multi-agent exploration mechanism to formulate an exceptionally detailed implementation plan.
+
+这是整段模板的"主谓宾"：**告诉 Claude Code 要以什么姿态、为什么目标工作**。默认的"多代理探索 + 实施计划"适合**软件研发/规划型**任务，但对很多其他场景（内容审校、数据分析、文案创意、市场研究、合规审查……）就不贴切。
+
+**建议你根据自己的目标改写这一段**，比如：
+- **内容审校专家**："You are a senior content reviewer specializing in {领域}. Your goal is to identify factual inaccuracies, tone inconsistencies, and structural weaknesses in the provided material."
+- **竞品分析专家**："Conduct a rigorous competitive analysis for {产品类别}. Produce a comparison matrix, positioning insights, and strategic recommendations."
+- **文案创意专家**："Generate multiple creative copy variants for {场景}, each with distinct positioning, tone, and call-to-action strategy."
+
+### 3. 工作流步骤（1–5 项，**根据任务复杂度适当精简或扩展**）
+
+调研专家原模板是 5 步：**探索 → 汇总 → 审查 → 提交计划 → 执行**。这套流程自带"多代理并行 + 交叉审查 + 计划审批"三层保障，适合跨度大、代价高的任务。但对很多轻量任务**过重了**，建议：
+
+- **任务简单**（单点查询/小改动）：删掉多代理 agent 派遣、删掉并行审查，直接 1 步"给出结论"。
+- **任务中等**：保留"探索 → 综合 → 审查"三步，删掉 ExitPlanMode 环节，直接产出结果。
+- **任务复杂、代价高**（大重构、多方案对比、跨多个领域调研）：保留 5 步，甚至增加"风险建模"或"方案对比矩阵"步骤。
+
+### 4. Step 1 的子角色列表（**按领域裁剪**）
+
+调研专家派了 6 种潜在角色（industry 调研员、学术调研员、综合 + 核实、竞品、Demo 产出、扩展角色）。**根据你的场景重新列清单**：
+- 写作类：可能需要"资料收集员 + 风格分析员 + 事实核查员"
+- 数据分析：可能需要"数据清洗员 + 统计建模员 + 可视化员"
+- 代码审计：可能需要"静态扫描员 + 依赖链审计员 + 威胁建模员"
+
+### 5. 最终产出清单（**对齐你的真正诉求**）
+
+> Your final plan must include the following elements: ...
+
+原模板要求的是"实施计划"的 6 个要素，但你的任务产出可能是：
+- 一份研究报告 → 改为"摘要 / 方法论 / 核心发现 / 限制 / 行动建议"
+- 一份审校报告 → 改为"问题清单 / 严重度分级 / 修改建议 / 改后示例"
+- 一份对比矩阵 → 改为"维度定义 / 打分标准 / 结论 / 优选理由"
+
+---
+
+## 写作建议（TL;DR）
+
+1. **保留外壳**（`<system-reminder>` 和 `[SCOPED INSTRUCTION]` 那行 ccv 自动加，你别重复写）。
+2. **重点改首段**：用一句话明确角色、目标、产出形式。
+3. **流程视复杂度伸缩**：轻量任务 1–2 步够了，复杂任务才走 5 步的多代理 + 审查闭环。
+4. **Step 1 的子角色按领域重写**：默认那些（academic papers / competitor / demo）大概率不是你要的。
+5. **最后的"产出清单"是 quality bar**：写清楚你期望的输出结构，Claude Code 会严格按这个出结果。
+
+---
+
+## 一个改好的例子（竞品分析专家）
+
+```
+You are a senior competitive intelligence analyst for {行业}. Your goal is to
+produce a decision-grade competitive landscape report for the product "{我方产品}".
+
+Instructions:
+1. Use the Agent tool to dispatch 3 parallel investigators:
+   - Market landscape agent: map the top 5–8 competitors with core positioning
+   - Feature matrix agent: compile a feature-by-feature comparison using
+     publicly available sources (webSearch)
+   - Pricing & GTM agent: analyze pricing models, distribution channels, and
+     go-to-market motions
+
+2. Synthesize the three streams into a unified competitive report.
+
+3. Dispatch one review agent to stress-test the report: challenge any
+   assumption lacking a cited source, flag outdated data (>12 months), and
+   propose one "non-obvious" insight.
+
+4. Deliver the final report with the following sections:
+   - TL;DR (3 bullets)
+   - Competitor positioning map
+   - Feature matrix (markdown table)
+   - Pricing & GTM table
+   - Top 3 strategic implications for our product
+   - Caveats & data gaps
+```
+
+这份相比调研专家原版：精简到 4 步，子角色从 6 个减到 3 个，产出清单完全重写成"报告章节"。