litclaude-ai 0.3.26 → 0.3.28

package/CHANGELOG.md

@@ -1,5 +1,24 @@
 # Changelog
 
+## 0.3.28 - 2026-07-14 — explicit start-work activation
+
+- Require the `$start-work` shorthand and natural `lit start work` activation
+  to be leading command phrases without changing other dollar routes.
+- Keep diagnostic prose, copied blockquotes, and explanatory mentions of
+  `$start-work` or `lit start work` inert while preserving explicit dollar,
+  bare, natural, and native slash start-work routes.
+
+## 0.3.27 - 2026-07-12 — structured interactive installer
+
+- Present interactive installs as five persistent, numbered Claude-native
+  stages with purpose and target details plus a live spinner during filesystem
+  work.
+- Add explicit `PREPARING INSTALL`, successful `INSTALL RECEIPT`, and failed
+  `INSTALL STOPPED` summaries without changing dry-run or deterministic
+  non-TTY `INSTALL_STEP` behavior.
+- Respect `NO_COLOR`, retain CI-safe output, and verify the installed payload,
+  Claude plugin registry, and compatibility pointer before reporting success.
+
 ## 0.3.26 - 2026-07-11 — planning quality and orchestration safety
 
 - Make `lit-plan` checklist depth adaptive: one bounded objective, explicit

package/README.md

@@ -9,7 +9,7 @@
 </p>
 <p align="center">
   <img src="https://img.shields.io/badge/npm-litclaude--ai-cb3837" />
-  <img src="https://img.shields.io/badge/version-0.3.26-2ea44f" />
+  <img src="https://img.shields.io/badge/version-0.3.28-2ea44f" />
   <img src="https://img.shields.io/badge/Claude%20Code-plugin-blueviolet" />
   <img src="https://img.shields.io/badge/license-MIT-blue" />
 </p>
@@ -22,10 +22,10 @@
 > `litclaude@litclaude-ai`, so normal `claude` launches can load the
 > LitClaude skills and hooks without a long `--plugin-dir` command.
 
-This checkout is prepared as `litclaude-ai@0.3.26` for personal install
+This checkout is prepared as `litclaude-ai@0.3.28` for personal install
 convenience. The repo can remain quiet; preparing npm package metadata here does
 not imply public repo promotion, marketplace publication, or advertisement.
-Future package releases still require explicit user approval. The v0.3.26 release material adds adaptive objective-achievable planning, draft-plan review, and orchestration readiness hardening. The v0.3.25 release material aligns portable details, public-read, litgoal status JSON, and evidence wording. The v0.3.24 release material aligns auxiliary skill inventories and advisory probes. The v0.3.23
+Future package releases still require explicit user approval. The v0.3.28 release material requires explicit leading start-work invocations so diagnostic or copied mentions remain inert. The v0.3.27 release material adds a structured, animated five-stage installer with a success or failure receipt while preserving deterministic automation output. The v0.3.26 release material adds adaptive objective-achievable planning, draft-plan review, and orchestration readiness hardening. The v0.3.25 release material aligns portable details, public-read, litgoal status JSON, and evidence wording. The v0.3.24 release material aligns auxiliary skill inventories and advisory probes. The v0.3.23
 release materials inject the bundled skill bodies for bare `hyperplan`,
 `litresearch`, `lit research`, `init-deep`, and explicit `$start-work` prompt-hook routes while
 preserving the v0.3.19 adversarial planning skill, the v0.3.18 read-only `lit-recap` session recap surface, the v0.3.17 separate-worker native `/goal` launcher, the v0.3.16
@@ -107,9 +107,14 @@ npx --yes litclaude-ai install
 
 During an interactive `npx` install, LitClaude shows every HUD brand color as
 a vivid live terminal preview and writes the selected accent into the managed
-Claude `statusLine` command. Animated installs also use colored setup, plugin,
-marketplace, registry, and cache labels so progress is easier to scan.
-Noninteractive installs keep deterministic plain `INSTALL_STEP` lines.
+Claude `statusLine` command. The installer then presents `PREPARING INSTALL`,
+five persistent numbered stages with each stage's purpose and target, an
+animated spinner while filesystem work is active, and an `INSTALL RECEIPT`
+that identifies the installed plugin, HUD, permission preference, and launch
+command. A failed stage prints `INSTALL STOPPED` and never reports success.
+`NO_COLOR` retains the structure without ANSI color. CI and non-TTY installs
+keep deterministic plain `INSTALL_STEP` lines, and `--dry-run` remains a
+side-effect-free preview.
 
 If you are currently inside this repository checkout, which has the same
 package name as the registry package, older published builds such as
@@ -120,7 +125,7 @@ directory, the normal install command works:
 
 ```bash
 cd /tmp
-npx --yes litclaude-ai@0.3.26 install
+npx --yes litclaude-ai@0.3.28 install
 ```
 
 Validate the installed plugin:
@@ -133,7 +138,7 @@ The installer also sets Claude Code's `statusLine` command to the packaged
 LitClaude HUD. A typical no-color render starts like:
 
 ```text
-[🔥LITCLAUDE v0.3.26] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
+[🔥LITCLAUDE v0.3.28] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
 ```
 
 The `↻` suffix is a compact rate-limit reset countdown. It is separated from
@@ -250,7 +255,9 @@ a visible hook message. Slash commands such as `/lit-plan` are handled by Claude
 Code's native command surface and should not double-activate the hook. Natural
 `lit start work` is a safe handoff: it emits `BLOCKED:` and tells the user to
 run `/start-work` or `/litclaude:start-work` with the approved plan because a
-prompt hook cannot switch Claude Code agents. Plain `lit` still maps to
+prompt hook cannot switch Claude Code agents. The `$start-work` shorthand and
+natural start-work phrases activate only at the beginning of the prompt, so diagnostic
+prose and copied or blockquoted mentions stay inert. Plain `lit` still maps to
 `/litclaude:lit-loop` / `Skill(lit-loop)` semantics before ordinary task
 execution, then the matching skill and agent instructions steer Claude toward
 test-first work, native goal handling, manual QA evidence, cleanup receipts, and

package/README_ko-KR.md

@@ -9,7 +9,7 @@
 </p>
 <p align="center">
   <img src="https://img.shields.io/badge/npm-litclaude--ai-cb3837" />
-  <img src="https://img.shields.io/badge/version-0.3.26-2ea44f" />
+  <img src="https://img.shields.io/badge/version-0.3.28-2ea44f" />
   <img src="https://img.shields.io/badge/Claude%20Code-plugin-blueviolet" />
   <img src="https://img.shields.io/badge/license-MIT-blue" />
 </p>
@@ -26,11 +26,11 @@
 > 설치되므로, 매번 긴 `--plugin-dir` 없이 일반 `claude` 실행에서
 > LitClaude skill과 hook을 불러올 수 있습니다.
 
-현재 checkout은 `litclaude-ai@0.3.26` 배포 준비용으로 정리되어 있습니다. 목적은
+현재 checkout은 `litclaude-ai@0.3.28` 배포 준비용으로 정리되어 있습니다. 목적은
 다른 PC에서도 빠르게 설치하기 위한 개인용 package metadata를 갖추는 것입니다.
 npm package metadata를 준비했다고 해서 홍보, 공개 저장소 운영, Claude
 marketplace 등록을 의미하지는 않습니다. 새 버전 배포는 항상 별도의 명시적
-승인 후에 진행합니다. v0.3.26 release material은 adaptive objective-achievable planning, draft-plan review, orchestration readiness hardening을 추가합니다. v0.3.25 release material은 portable details, public-read, litgoal status JSON, evidence wording을 정렬합니다. v0.3.24 release material은 auxiliary skill inventory와 advisory probe를 정렬합니다. v0.3.23 release material은 bare `hyperplan`,
+승인 후에 진행합니다. v0.3.28 release material은 start-work 호출을 prompt 선두의 명시적 invocation으로 제한해 진단문이나 복사된 언급을 비활성 상태로 유지합니다. v0.3.27 release material은 success/failure receipt와 deterministic automation output을 갖춘 5-stage animated installer를 추가합니다. v0.3.26 release material은 adaptive objective-achievable planning, draft-plan review, orchestration readiness hardening을 추가합니다. v0.3.25 release material은 portable details, public-read, litgoal status JSON, evidence wording을 정렬합니다. v0.3.24 release material은 auxiliary skill inventory와 advisory probe를 정렬합니다. v0.3.23 release material은 bare `hyperplan`,
 `litresearch`, `lit research`, `init-deep`, explicit `$start-work` prompt-hook route에
 bundled skill body를 주입하면서, v0.3.19 adversarial planning
 skill, v0.3.18 read-only `lit-recap` session recap surface, v0.3.17 별도 worker 기반 native `/goal`
@@ -110,10 +110,12 @@ npx --yes litclaude-ai install
 
 interactive `npx` 설치에서는 LitClaude가 HUD brand color 후보를 더 선명한
 실제 터미널 색상으로 바로 보여주고, 선택한 accent를 Claude의 managed
-`statusLine` command에 저장합니다. animated install에서는 setup, plugin,
-marketplace, registry, cache label도 색상으로 구분해서 진행 상황을 더 쉽게
-스캔할 수 있습니다. noninteractive install은 deterministic한 plain
-`INSTALL_STEP` line을 유지합니다.
+`statusLine` command에 저장합니다. 이후 `PREPARING INSTALL`, 목적과 target을
+표시하는 5개의 numbered stage, 파일시스템 작업 중 animated spinner, 그리고
+plugin·HUD·permission·launch command를 정리한 `INSTALL RECEIPT`를 보여줍니다.
+중간 단계가 실패하면 `INSTALL STOPPED`를 출력하고 성공을 보고하지 않습니다.
+`NO_COLOR`는 ANSI 색상만 제거하고 구조를 유지합니다. CI/non-TTY 설치는 deterministic한
+plain `INSTALL_STEP` line을 유지하며 `--dry-run`은 계속 side-effect-free preview입니다.
 
 현재 위치가 이 저장소 checkout이면, registry package와 같은 package name을
 가진 same-name source checkout 안에 있는 상태입니다. 이 경우
@@ -124,7 +126,7 @@ checkout을 먼저 해석해서 `sh: litclaude-ai: command not found`로 실패
 
 ```bash
 cd /tmp
-npx --yes litclaude-ai@0.3.26 install
+npx --yes litclaude-ai@0.3.28 install
 ```
 
 설치 상태를 확인합니다.
@@ -137,7 +139,7 @@ installer는 Claude Code의 `statusLine` command도 packaged LitClaude HUD로
 설정합니다. 색상을 제거한 예시는 다음처럼 시작합니다.
 
 ```text
-[🔥LITCLAUDE v0.3.26] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
+[🔥LITCLAUDE v0.3.28] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
 ```
 
 `↻` 표시는 rate-limit reset까지 남은 시간을 짧게 보여주는 countdown입니다.
@@ -246,8 +248,9 @@ visible hook message를 보여주는 것입니다. `/lit-plan` 같은 slash comm
 Claude Code native command surface에서 처리되며 hook을 중복 활성화하지
 않습니다. 자연어 `lit start work`는 안전한 handoff로 `BLOCKED:`를 내고,
 prompt hook이 Claude Code agent를 전환할 수 없으므로 승인된 plan과 함께
-`/start-work` 또는 `/litclaude:start-work`를 실행하라고 안내합니다. plain
-`lit`는 일반 작업 실행 전 `/litclaude:lit-loop` / `Skill(lit-loop)` semantics로
+`/start-work` 또는 `/litclaude:start-work`를 실행하라고 안내합니다.
+`$start-work`와 자연어 start-work phrase는 prompt 선두에서만 활성화되므로 진단문,
+복사문, blockquote 안의 단순 언급은 활성화되지 않습니다. plain `lit`는 일반 작업 실행 전 `/litclaude:lit-loop` / `Skill(lit-loop)` semantics로
 처리됩니다. 그 뒤 skill 및 agent 지시문이 Claude를 test-first 작업, native
 goal 처리, 실제 수동 QA 증거, cleanup receipt, 향후 release step의 명시적 승인
 쪽으로 유도합니다.

package/RELEASE_CHECKLIST.md

@@ -1,6 +1,8 @@
 # LitClaude Release Checklist
 
-Status: `litclaude-ai@0.3.26` is the current release candidate — an adaptive
+Status: `litclaude-ai@0.3.28` is the current release candidate — explicit
+leading start-work activation guards plus a structured five-stage interactive
+installer with success/failure receipts and adaptive
 planning, draft-plan review, and orchestration-safety release on top of the portable
 details, public-read, litgoal-status, and evidence-wording release and the auxiliary
 skill inventory and advisory-probe release and the full bare
@@ -15,9 +17,9 @@ side-effect-free, the launcher starts only a separate Claude Code
 print/background worker, and the release preserves the Korean polishing
 command, strict multi-agent review pipeline, fidelity guardrails, package
 hygiene checks, native route gates, and safe start-work handoff behavior.
-`package.json` is aligned to `0.3.26`,
-`plugins/litclaude/.claude-plugin/plugin.json` is aligned to `0.3.26`, and the
-plugin-local MCP server reports `0.3.26`.
+`package.json` is aligned to `0.3.28`,
+`plugins/litclaude/.claude-plugin/plugin.json` is aligned to `0.3.28`, and the
+plugin-local MCP server reports `0.3.28`.
 
 This release carries the v0.2.2 Dynamic workflow hardening surfaces:
 `/dynamic-workflow`, `workflow-check --json`, native `/goal` fallback guidance,
@@ -66,10 +68,10 @@ No npm publication is required for this track.
 Before requesting publication approval, confirm these artifacts from the current
 checkout:
 
-- `package.json` version is `0.3.26`.
-- `plugins/litclaude/.claude-plugin/plugin.json` version is `0.3.26`.
-- `plugins/litclaude/bin/litclaude-mcp.js` reports server version `0.3.26`.
-- Prompt-hook tests cover bundled `SKILL.md` body injection for bare `hyperplan`, `litresearch`, `lit research`, `init-deep`, and explicit `$start-work`, while natural-language `lit start work` stays BLOCKED.
+- `package.json` version is `0.3.28`.
+- `plugins/litclaude/.claude-plugin/plugin.json` version is `0.3.28`.
+- `plugins/litclaude/bin/litclaude-mcp.js` reports server version `0.3.28`.
+- Prompt-hook tests cover bundled `SKILL.md` body injection for bare `hyperplan`, `litresearch`, `lit research`, `init-deep`, and explicit leading `$start-work`; diagnostic/copy mentions stay inert while leading natural-language `lit start work` stays BLOCKED.
 - `lit search` and `lit query` route to `/litclaude:litresearch` without activating on slash mentions, code spans, or non-lit prompts.
 - Litresearch web lanes require public API/feed preference, validator-first checks, route traces, prompt-injection quarantine, and honest auth/paywall/private-data stop reasons.
 - `node bin/litclaude-ai.js public-read <public-url> --json` exposes the guarded JS runtime reader.

package/bin/litclaude-ai.js

@@ -10,6 +10,7 @@ import {
   symlinkSync,
   writeFileSync,
 } from "node:fs";
+import { cp, mkdir, rm } from "node:fs/promises";
 import { homedir } from "node:os";
 import { dirname, join, resolve } from "node:path";
 import { createInterface } from "node:readline/promises";
@@ -169,6 +170,8 @@ const shouldAnimateInstall = () => {
   return Boolean(process.stdout.isTTY && !process.env.CI);
 };
 
+const shouldColorInstall = () => !Object.hasOwn(process.env, "NO_COLOR");
+
 const installStepTheme = (label) => {
   if (/registry|HUD/iu.test(label)) return { code: 215, tag: "registry" };
   if (/plugin/iu.test(label)) return { code: 141, tag: "plugin" };
@@ -177,32 +180,87 @@ const installStepTheme = (label) => {
   return { code: 81, tag: "setup" };
 };
 
-const color256 = (code, text) => `\x1b[38;5;${code}m${text}\x1b[0m`;
+const color256 = (code, text) => shouldColorInstall() ? `\x1b[38;5;${code}m${text}\x1b[0m` : text;
+const bold = (text) => shouldColorInstall() ? `\x1b[1m${text}\x1b[0m` : text;
 
 const formatInstallStep = (label) => {
   const theme = installStepTheme(label);
-  return `${color256(theme.code, theme.tag)} \x1b[1m${label}\x1b[0m`;
+  return `${color256(theme.code, theme.tag)} ${bold(label)}`;
 };
 
 const createInstallProgress = () => {
   const animated = shouldAnimateInstall();
   let frameIndex = 0;
+  const live = animated && Boolean(process.stdout.isTTY);
+
+  const clearLiveLine = () => {
+    if (live) process.stdout.write("\r\x1b[2K");
+  };
+
   return {
-    step(label) {
+    prepare({ claudeTarget, litTarget }) {
       if (!animated) {
-        process.stdout.write(`INSTALL_STEP: ${label}\n`);
+        process.stdout.write(`INSTALL_STEP: Installing LitClaude ${version}\n`);
         return;
       }
-      const frame = spinnerFrames[frameIndex % spinnerFrames.length];
-      frameIndex += 1;
-      process.stdout.write(`${color256(81, frame)} ${formatInstallStep(label)}\n`);
+      process.stdout.write(
+        `\n╭─ PREPARING INSTALL\n` +
+        `│ Package       LitClaude ${version}\n` +
+        `│ Action        Installing LitClaude ${version}\n` +
+        `│ Claude state  ${claudeTarget}\n` +
+        `│ Local state   ${litTarget}\n` +
+        `│ Progress      5 persistent stages with a final verification\n` +
+        `╰─ Existing unrelated Claude settings are preserved\n\n`,
+      );
     },
-    done(label) {
+    async run({ number, title, label, purpose, target }, work) {
       if (!animated) {
         process.stdout.write(`INSTALL_STEP: ${label}\n`);
-        return;
+        return work();
+      }
+
+      process.stdout.write(
+        `${String(number).padStart(2, "0")} / 05  ${title}\n` +
+        `     ${purpose}\n` +
+        `     target  ${target}\n`,
+      );
+
+      const renderSpinner = () => {
+        const frame = spinnerFrames[frameIndex % spinnerFrames.length];
+        frameIndex += 1;
+        const line = `  ${color256(81, frame)} ${formatInstallStep(label)}`;
+        process.stdout.write(live ? `\r\x1b[2K${line}` : `${line}\n`);
+      };
+      renderSpinner();
+      const timer = live ? setInterval(renderSpinner, 80) : undefined;
+
+      try {
+        const result = await work();
+        if (timer) clearInterval(timer);
+        clearLiveLine();
+        process.stdout.write(`  ${color256(119, "✓")} ${bold(label)}\n\n`);
+        return result;
+      } catch (error) {
+        if (timer) clearInterval(timer);
+        clearLiveLine();
+        process.stdout.write(
+          `  ${color256(203, "✗")} ${bold(`${label} failed`)}\n\n` +
+          `╭─ INSTALL STOPPED\n` +
+          `│ Stage         ${String(number).padStart(2, "0")} / 05 · ${title}\n` +
+          `│ Reason        ${error instanceof Error ? error.message : String(error)}\n` +
+          `│ Next          Resolve the reason above, then rerun litclaude install\n` +
+          `╰─ Exit without reporting success\n`,
+        );
+        throw error;
       }
-      process.stdout.write(`${color256(119, "✔")} \x1b[1m${label}\x1b[0m\n`);
+    },
+    receipt(rows) {
+      if (!animated) return;
+      process.stdout.write("\n╭─ INSTALL RECEIPT\n");
+      for (const [key, value] of rows) {
+        process.stdout.write(`│ ${key.padEnd(13)} ${value}\n`);
+      }
+      process.stdout.write("╰─ Installation complete\n\n");
     },
   };
 };
@@ -560,6 +618,20 @@ const resetCurrentPointer = (home, target) => {
   }
 };
 
+const verifyInstalledPayload = ({ home, litClaudeHome, targetPlugin }) => {
+  for (const relativePath of [".claude-plugin/plugin.json", "hooks/hooks.json", "bin/litclaude-hook.js"]) {
+    if (!existsSync(join(targetPlugin, relativePath))) {
+      throw new Error(`installed plugin is missing ${relativePath}`);
+    }
+  }
+  if (!readInstalledPlugins(home).plugins[pluginKey]?.length) {
+    throw new Error(`Claude registry is missing ${pluginKey}`);
+  }
+  if (!existsSync(currentRoot(litClaudeHome))) {
+    throw new Error("LitClaude compatibility pointer was not created");
+  }
+};
+
 const install = async ({ dryRun, rest }) => {
   const { permissionMode, permissionExplicit } = parseInstallOptions(rest);
   const home = claudeHome();
@@ -586,24 +658,78 @@ const install = async ({ dryRun, rest }) => {
   }
 
   const hudAccent = await chooseHudAccent();
-  progress.step(`Installing LitClaude ${version}`);
-  progress.step("Plugin payload: copying Claude Code plugin");
-  rmSync(targetPlugin, { recursive: true, force: true });
-  mkdirSync(dirname(targetPlugin), { recursive: true });
-  cpSync(sourcePlugin, targetPlugin, { recursive: true });
-  progress.step("Marketplace: writing local Claude marketplace");
-  const localMarketplace = writeLocalMarketplace(litClaudeHome);
-  progress.step("Claude registry: enabling plugin and HUD");
-  registerClaudePlugin(targetPlugin, localMarketplace, home, hudAccent, permissionMode);
+  progress.prepare({ claudeTarget: home, litTarget: litClaudeHome });
+  await progress.run(
+    {
+      number: 1,
+      title: "PLUGIN PAYLOAD",
+      label: "Plugin payload copied",
+      purpose: "Copy the packaged hooks, skills, agents, MCP, and LSP surfaces",
+      target: targetPlugin,
+    },
+    async () => {
+      await rm(targetPlugin, { recursive: true, force: true });
+      await mkdir(dirname(targetPlugin), { recursive: true });
+      await cp(sourcePlugin, targetPlugin, { recursive: true });
+    },
+  );
+  const localMarketplace = await progress.run(
+    {
+      number: 2,
+      title: "LOCAL MARKETPLACE",
+      label: "Marketplace metadata written",
+      purpose: "Create a local Claude marketplace without a GitHub login",
+      target: targetMarketplace,
+    },
+    () => writeLocalMarketplace(litClaudeHome),
+  );
+  await progress.run(
+    {
+      number: 3,
+      title: "CLAUDE REGISTRY",
+      label: "Claude plugin and HUD enabled",
+      purpose: "Enable the plugin, statusLine HUD, and permission preference",
+      target: claudeSettingsPath(home),
+    },
+    () => registerClaudePlugin(targetPlugin, localMarketplace, home, hudAccent, permissionMode),
+  );
 
-  progress.step("Compatibility cache: refreshing local current pointer");
   const litRoot = versionRoot(litClaudeHome);
-  rmSync(litRoot, { recursive: true, force: true });
-  mkdirSync(dirname(pluginPathForRoot(litRoot)), { recursive: true });
-  cpSync(sourcePlugin, pluginPathForRoot(litRoot), { recursive: true });
-  resetCurrentPointer(litClaudeHome, litRoot);
+  await progress.run(
+    {
+      number: 4,
+      title: "COMPATIBILITY CACHE",
+      label: "Compatibility pointer refreshed",
+      purpose: "Refresh the versioned fallback used by LitClaude helper commands",
+      target: currentRoot(litClaudeHome),
+    },
+    async () => {
+      await rm(litRoot, { recursive: true, force: true });
+      await mkdir(dirname(pluginPathForRoot(litRoot)), { recursive: true });
+      await cp(sourcePlugin, pluginPathForRoot(litRoot), { recursive: true });
+      resetCurrentPointer(litClaudeHome, litRoot);
+    },
+  );
+  await progress.run(
+    {
+      number: 5,
+      title: "INSTALL VERIFICATION",
+      label: "Installed surfaces verified",
+      purpose: "Confirm the payload, Claude registry, and compatibility pointer",
+      target: targetPlugin,
+    },
+    () => verifyInstalledPayload({ home, litClaudeHome, targetPlugin }),
+  );
 
-  progress.done("LitClaude install complete");
+  progress.receipt([
+    ["Status", "Ready for Claude Code"],
+    ["Version", version],
+    ["Plugin", pluginKey],
+    ["Plugin path", intendedPluginPath(home)],
+    ["HUD", `${hudAccent} accent`],
+    ["Permissions", permissionMode],
+    ["Launch", "claude"],
+  ]);
   process.stdout.write(`INSTALL_PASS: LitClaude ${version} installed\n`);
   process.stdout.write(`Claude plugin: ${pluginKey}\n`);
   process.stdout.write(`Marketplace: ${localMarketplace}\n`);

package/docs/hooks.md

@@ -121,7 +121,9 @@ surfaces and do not activate this prompt hook; code spans, code fences,
 substrings, and compound tokens are also ignored. Natural `lit start work`
 returns a `BLOCKED:` handoff that tells the user to run `/start-work` or
 `/litclaude:start-work` with the approved plan, because a prompt hook cannot
-switch Claude Code agents.
+switch Claude Code agents. The `$start-work` shorthand and natural start-work
+phrases activate only when they lead the prompt; diagnostic prose and copied or
+blockquoted mentions remain inert.
 
 LitClaude follows the LitClaude goal pattern as an honest native-binding attempt,
 not as slash-command injection. If Claude Code exposes native goal tools, the

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "litclaude-ai",
-  "version": "0.3.26",
+  "version": "0.3.28",
   "description": "Claude Code-native workflow distribution.",
   "type": "module",
   "bin": {

package/plugins/litclaude/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "litclaude",
   "description": "Claude Code-native workflow plugin.",
-  "version": "0.3.26",
+  "version": "0.3.28",
   "author": {
     "name": "LitClaude contributors"
   },

package/plugins/litclaude/bin/litclaude-hook.js

@@ -159,6 +159,7 @@ const containsSlashCommandMention = (text) => slashCommandMentionPattern.test(te
 const findDollarCommandTrigger = (raw) => {
   const match = /(?:^|\s)\$(deep-interview|dynamic-workflow|init-deep|lit-plan|lit-recap|lit-loop|start-work|review-work|litgoal|litresearch)(?=$|[^\w-])/u.exec(raw);
   if (!match) return undefined;
+  if (match[1] === "start-work" && !/^\s*\$start-work(?=$|[^\w-])/u.test(raw)) return undefined;
   return { ...triggerByToken[match[1]], source: "dollar-command" };
 };
 
@@ -169,6 +170,7 @@ const findLeadingBareCommandTrigger = (raw) => {
 };
 
 const hasLitTrigger = (raw) => containsCompoundBoundedWord(raw, "lit") || containsCompoundBoundedWord(raw, "litwork");
+const hasTrailingLitInvocation = (raw) => /(?:^|\s)lit(?:work)?\s*$/u.test(raw);
 
 const hangulSyllable = /^[가-힣]$/u;
 
@@ -205,12 +207,13 @@ const findNaturalLitTrigger = (prompt) => {
   if (hasExplicitTeamRoute(normalized)) return { ...triggerByToken["agent-team"], source: "natural-language" };
   if (hasExplicitWorkflowRoute(normalized)) return { ...triggerByToken["dynamic-workflow"], source: "natural-language" };
   if (containsPlainWord(normalized, "review")) return { ...triggerByToken["review-work"], source: "natural-language" };
-  if (containsPlainWord(normalized, "start") && containsPlainWord(normalized, "work")) {
+  if (/^\s*lit(?:work)?\s+start\s+work(?=$|[^\w-])/u.test(normalized)) {
     return { ...triggerByToken["start-work"], source: "natural-language", safetyBlock: true };
   }
   if (hasExplicitResearchRoute(normalized)) return { ...triggerByToken.litresearch, source: "natural-language" };
   if (containsPlainWord(normalized, "goal")) return { ...triggerByToken.litgoal, source: "natural-language" };
   if (containsPlainWord(normalized, "plan")) return { ...triggerByToken["lit-plan"], source: "natural-language" };
+  if (containsPlainWord(normalized, "start") && containsPlainWord(normalized, "work") && !hasTrailingLitInvocation(normalized)) return undefined;
   return { ...triggerByToken["lit-loop"], source: "natural-language", softConfirm: true };
 };
 
@@ -218,6 +221,7 @@ const findWorkflowTrigger = (prompt) => {
   const raw = rawTriggerText(prompt);
   const dollarTrigger = findDollarCommandTrigger(raw);
   if (dollarTrigger) return dollarTrigger;
+  if (/(?:^|\s)\$start-work(?=$|[^\w-])/u.test(raw) && !hasTrailingLitInvocation(raw)) return undefined;
   if (containsSlashCommandMention(raw)) return undefined;
   const bareCommandTrigger = findLeadingBareCommandTrigger(raw);
   if (bareCommandTrigger) return bareCommandTrigger;

package/plugins/litclaude/bin/litclaude-mcp.js

@@ -3,7 +3,7 @@
 import { readPublicSource } from "../lib/public-source-reader/reader.mjs";
 
 const protocolVersion = "2024-11-05";
-const serverVersion = "0.3.26";
+const serverVersion = "0.3.28";
 
 const publicSourceReadTool = {
   name: "public_source_read",