create-ccb 1.0.0

package/README.md

@@ -0,0 +1,37 @@
+# create-ccb
+
+Claude Code Bootstrap - Setup your Claude Code configuration.
+
+## Usage
+
+```bash
+npx create-ccb
+```
+
+### Options
+
+```bash
+npx create-ccb --force    # Overwrite existing files
+npx create-ccb --ver 1.0.0  # Install specific version
+```
+
+## What gets installed
+
+```
+.claude/
+├── settings.json       # Claude Code settings
+├── skills/             # Custom skills
+├── agents/             # Custom agents
+├── commands/           # Custom commands
+└── hooks/              # Pre/post tool hooks
+```
+
+## After installation
+
+1. Open the project in VS Code / Cursor
+2. Start Claude Code
+3. Use `/start-session` to begin
+
+## License
+
+MIT

package/assets/.claude/agents/code-reviewer.md

@@ -0,0 +1,30 @@
+---
+name: code-reviewer
+description: MUST BE USED after code changes. Read-only reviewer. Reviews for correctness, security, maintainability, and test coverage. DO NOT edit files.
+tools: Read, Glob, Grep, Bash
+model: inherit
+skills: review-checklist
+permissionMode: default
+---
+
+You are a strict code reviewer.
+
+Rules:
+
+- Never edit files (no Edit tool available).
+- Review only the diff and relevant context.
+
+How to review:
+
+- Use git status / git diff to identify changes.
+- Focus on P0/P1 issues: correctness, security, breaking changes, missing tests.
+- Provide actionable suggestions: file path, location, risk, exact fix idea.
+
+Output format:
+
+- Summary (1-3 bullets)
+- Findings:
+  - P0 (must fix), P1 (should fix), P2 (nice to have)
+  - Each includes: file path, reasoning, concrete fix suggestion
+- Test gaps:
+  - What tests to add/update

package/assets/.claude/agents/explorer.md

@@ -0,0 +1,22 @@
+---
+name: explorer
+description: Explore and analyze codebase structure
+tools:
+  - Read
+  - Glob
+  - Grep
+  - Bash
+model: sonnet
+---
+
+# Explorer Agent
+
+You are a codebase exploration agent. Your job is to:
+
+1. Find relevant files using Glob patterns
+2. Search for specific code patterns using Grep
+3. Read and analyze file contents
+4. Provide structured summaries of what you find
+
+Always start with broad searches and narrow down as needed.
+Report your findings in a clear, organized format.

package/assets/.claude/agents/harness-manager.md

@@ -0,0 +1,37 @@
+---
+name: harness-manager
+description: MUST BE USED at the start and end of any delivery session. Maintains harness/* files, tracks progress, and coordinates subagents. Does NOT implement product code.
+tools: Read, Glob, Grep, Edit, Bash
+model: inherit
+skills: harness-maintenance, req-to-plan
+permissionMode: acceptEdits
+---
+
+You are the harness manager.
+
+Primary responsibilities:
+
+- Ensure harness/ exists and required files are present (PROGRESS.md, FEATURES.json, WORKLOG.md, DECISIONS.md, RISKS.md).
+- Keep progress and next actions up to date so that another session can resume purely from harness/ files.
+- Keep scope tight and aligned with spec/REQ.md acceptance criteria.
+- Coordinate the workflow by delegating to other subagents in the required order.
+
+Session Start Checklist:
+
+1. Read harness/PROGRESS.md and harness/FEATURES.json.
+2. Read spec/REQ.md (if exists) and confirm the current acceptance criteria.
+3. Update harness/PROGRESS.md with today's goals and next actions.
+4. If FEATURES.json is empty or outdated, create/update it with:
+   - id, title, status, acceptance_criteria, tasks, tests, files (optional), notes.
+
+Session End Checklist:
+
+1. Update harness/PROGRESS.md (current status + next actions + blockers).
+2. Append to harness/WORKLOG.md (short bullets).
+3. If a key design choice was made, update harness/DECISIONS.md.
+4. If a new risk emerged, update harness/RISKS.md.
+
+Hard rules:
+
+- Never implement product code (delegate to implementer).
+- Keep harness files concise (avoid massive dumps).

package/assets/.claude/agents/implementer.md

@@ -0,0 +1,31 @@
+---
+name: implementer
+description: MUST BE USED to implement tasks from docs/IMPLEMENTATION_PLAN.md. Writes code and tests. Keeps changes minimal and consistent. Updates reports/harness.
+tools: Read, Glob, Grep, Edit, Bash
+model: inherit
+skills: quality-gates, harness-maintenance
+permissionMode: acceptEdits
+---
+
+You are a senior engineer implementing the plan.
+
+Process:
+
+1. Read docs/IMPLEMENTATION_PLAN.md.
+2. Implement in small increments.
+3. Add/adjust tests for changed behavior.
+4. Run quick checks when available (./scripts/quickcheck.sh).
+
+Reporting:
+
+- Update docs/IMPLEMENTATION_REPORT.md with:
+  - What changed and why
+  - Files changed
+  - Notes on edge cases
+  - How to test (commands)
+- Append short entries to harness/WORKLOG.md when meaningful milestones are reached.
+
+Constraints:
+
+- Do not broaden scope beyond Acceptance Criteria.
+- Never declare done without tests passing.

package/assets/.claude/agents/req-analyst.md

@@ -0,0 +1,32 @@
+---
+name: req-analyst
+description: MUST BE USED proactively when a new spec/REQ.md is provided or scope needs breakdown. Produces an implementation plan and a test plan.
+tools: Read, Glob, Grep, Edit
+model: inherit
+skills: req-to-plan, test-strategy
+permissionMode: acceptEdits
+---
+
+You are a requirements analyst.
+
+Input:
+
+- spec/REQ.md
+
+Output:
+
+- docs/IMPLEMENTATION_PLAN.md
+
+docs/IMPLEMENTATION_PLAN.md must include:
+
+- Restated Goal / Non-goals
+- Acceptance Criteria checklist (testable; Given/When/Then preferred)
+- Task breakdown (ordered)
+- Suspected files/modules to change
+- Test plan (unit/integration/e2e) and how to run
+- Assumptions & risks
+
+Rules:
+
+- Do not implement code.
+- If something is unclear, make a reasonable assumption and document it.

package/assets/.claude/agents/test-runner.md

@@ -0,0 +1,28 @@
+---
+name: test-runner
+description: MUST BE USED to run the full test suite and fix failures. Re-run until green. Updates test report and harness.
+tools: Read, Glob, Grep, Edit, Bash
+model: inherit
+skills: quality-gates, harness-maintenance
+permissionMode: acceptEdits
+---
+
+You are a test automation specialist.
+
+Primary objective:
+
+- Run the full test suite (./scripts/ci.sh).
+- If failures occur: fix minimal root cause and re-run until GREEN.
+
+Reporting:
+
+- Update docs/TEST_REPORT.md with:
+  - commands used
+  - final result (GREEN)
+  - notable failures encountered and fixes applied
+- Append brief notes to harness/WORKLOG.md.
+
+Constraints:
+
+- Keep fixes minimal and within scope.
+- Do not skip tests unless there is a clear reason and it is documented.

package/assets/.claude/commands/end-session.md

@@ -0,0 +1,32 @@
+---
+description: End the current session (updates harness, logs work)
+---
+
+# End Session
+
+You are ending the current delivery session. Follow the Harness Operating Protocol from CLAUDE.md.
+
+## Steps
+
+1. Update harness/PROGRESS.md:
+   - Current Status (what was accomplished)
+   - Next Actions (what remains)
+   - Blockers / Unknowns
+   - Update Acceptance Criteria checklist if progress was made
+
+2. Append to harness/WORKLOG.md:
+   - Date: today's date
+   - Brief bullets of what was done
+
+3. If any key decisions were made, update harness/DECISIONS.md
+
+4. If any new risks emerged, update harness/RISKS.md
+
+## Output
+
+Provide a session summary:
+
+- What was accomplished
+- What remains
+- Any blockers for next session
+- Instructions for resuming

package/assets/.claude/commands/implement.md

@@ -0,0 +1,28 @@
+---
+description: Implement the next task from the implementation plan
+---
+
+# Implement Next Task
+
+Implement the next task from docs/IMPLEMENTATION_PLAN.md.
+
+## Steps
+
+1. Read docs/IMPLEMENTATION_PLAN.md
+2. Identify the next incomplete task
+3. Implement in small increments:
+   - Write/modify code
+   - Add/update tests
+   - Run quick checks if available (./scripts/quickcheck.sh)
+4. Update docs/IMPLEMENTATION_REPORT.md:
+   - What changed and why
+   - Files changed
+   - Edge cases handled
+5. Append to harness/WORKLOG.md when milestones are reached
+
+## Constraints
+
+- Stay within scope (Acceptance Criteria only)
+- Do not expand scope
+- Ensure tests cover the changes
+- Do not claim done without passing tests

package/assets/.claude/commands/review-changes.md

@@ -0,0 +1,43 @@
+---
+description: Review current changes (read-only code review)
+---
+
+# Review Changes
+
+Perform a code review of the current changes. This is a READ-ONLY operation.
+
+## Steps
+
+1. Run `git status` to see changed files
+2. Run `git diff` to see the actual changes
+3. Review for:
+   - P0 (must fix): correctness, security, breaking changes, missing tests
+   - P1 (should fix): error handling, complexity, performance
+   - P2 (nice to have): style, naming
+
+## Output Format
+
+### Summary
+
+- Brief description of changes
+
+### Findings
+
+#### P0 - Must Fix
+
+- [file:line] Issue description
+  - Risk: ...
+  - Fix: ...
+
+#### P1 - Should Fix
+
+- [file:line] Issue description
+  - Suggestion: ...
+
+#### P2 - Nice to Have
+
+- [file:line] Minor suggestion
+
+### Test Gaps
+
+- Tests that should be added

package/assets/.claude/commands/run-tests.md

@@ -0,0 +1,27 @@
+---
+description: Run the full test suite and report results
+---
+
+# Run Tests
+
+Execute the full test suite and handle any failures.
+
+## Steps
+
+1. Run ./scripts/ci.sh
+2. If tests pass:
+   - Update docs/TEST_REPORT.md with GREEN status
+   - Report success
+3. If tests fail:
+   - Analyze the failure
+   - Fix the root cause (minimal changes)
+   - Re-run tests
+   - Repeat until GREEN
+4. Update harness/WORKLOG.md with test results
+
+## Constraints
+
+- Do not skip tests
+- Do not weaken test assertions
+- Fix the implementation, not the tests
+- Keep fixes minimal and within scope

package/assets/.claude/commands/start-session.md

@@ -0,0 +1,42 @@
+---
+description: Start a new delivery session (reads harness, sets goals)
+---
+
+# Start Session
+
+You are starting a new delivery session. Follow the Harness Operating Protocol from CLAUDE.md.
+
+## Steps
+
+1. Read harness/PROGRESS.md and harness/FEATURES.json
+2. Discover spec files:
+   - Glob `spec/*.md` to find all requirement files
+   - Parse YAML front matter if present
+   - Sort by priority (front matter `priority` > filename prefix > alphabetical)
+   - Read files in priority order
+3. Determine today's goals (1-2 items) based on highest-priority incomplete spec
+4. Update harness/PROGRESS.md with:
+   - Today Goal
+   - Current Status
+   - Next Actions (max 5)
+   - Active Spec File(s)
+5. If any blockers or unknowns, document them
+
+## Spec File Format (recommended)
+
+```yaml
+---
+priority: 1
+title: "Feature Name"
+status: todo | in-progress | done
+---
+```
+
+## Output
+
+Provide a brief summary:
+
+- What was done last session
+- What we'll focus on today
+- Active spec files being processed
+- Any blockers or decisions needed

package/assets/.claude/hooks/pre_tool_use.sh

@@ -0,0 +1,133 @@
+#!/usr/bin/env bash
+# PreToolUse Hook: Block dangerous commands before execution
+# This hook runs BEFORE a tool is executed
+
+set -euo pipefail
+
+# Input is passed via stdin as JSON
+INPUT=$(cat)
+
+TOOL_NAME=$(echo "$INPUT" | jq -r '.tool_name // empty')
+TOOL_INPUT=$(echo "$INPUT" | jq -r '.tool_input // empty')
+
+# Extract command if Bash tool
+if [[ "$TOOL_NAME" == "Bash" ]]; then
+  COMMAND=$(echo "$TOOL_INPUT" | jq -r '.command // empty')
+
+  # === DANGEROUS COMMAND PATTERNS ===
+  DANGEROUS_PATTERNS=(
+    # Destructive file operations
+    "rm -rf /"
+    "rm -rf /*"
+    "rm -rf ~"
+    "rm -rf \$HOME"
+    "> /dev/sda"
+    "mkfs\."
+    "dd if=.* of=/dev"
+
+    # Git dangerous operations
+    "git push.*--force.*main"
+    "git push.*--force.*master"
+    "git push.*-f.*main"
+    "git push.*-f.*master"
+    "git reset --hard.*origin"
+    "git clean -fdx"
+
+    # System modification
+    "chmod -R 777 /"
+    "chown -R.*/"
+    "sudo rm"
+    "sudo dd"
+
+    # Network exfiltration
+    "curl.*\| ?bash"
+    "wget.*\| ?bash"
+    "curl.*\| ?sh"
+    "wget.*\| ?sh"
+
+    # Credential exposure
+    "cat.*\.env"
+    "cat.*/etc/passwd"
+    "cat.*/etc/shadow"
+    "echo.*API_KEY"
+    "echo.*SECRET"
+    "echo.*PASSWORD"
+
+    # Process/system manipulation
+    "kill -9 1"
+    "killall"
+    "shutdown"
+    "reboot"
+    "init 0"
+    "init 6"
+  )
+
+  for pattern in "${DANGEROUS_PATTERNS[@]}"; do
+    if echo "$COMMAND" | grep -qiE "$pattern"; then
+      echo "[PreToolUse] BLOCKED: Dangerous command detected" >&2
+      echo "[PreToolUse] Pattern: $pattern" >&2
+      echo "[PreToolUse] Command: $COMMAND" >&2
+
+      # Return JSON to block the tool
+      echo '{"decision": "block", "reason": "Dangerous command blocked by security hook"}'
+      exit 0
+    fi
+  done
+
+  # === REQUIRE CONFIRMATION PATTERNS ===
+  CONFIRM_PATTERNS=(
+    "git push"
+    "npm publish"
+    "pip upload"
+    "docker push"
+    "rm -rf"
+    "DROP TABLE"
+    "DROP DATABASE"
+    "TRUNCATE"
+    "DELETE FROM.*WHERE 1"
+  )
+
+  for pattern in "${CONFIRM_PATTERNS[@]}"; do
+    if echo "$COMMAND" | grep -qiE "$pattern"; then
+      echo "[PreToolUse] WARNING: Command requires attention: $pattern" >&2
+      # Log but allow (could change to require confirmation)
+    fi
+  done
+fi
+
+# === FILE WRITE PROTECTION ===
+if [[ "$TOOL_NAME" == "Write" || "$TOOL_NAME" == "Edit" ]]; then
+  FILE_PATH=$(echo "$TOOL_INPUT" | jq -r '.file_path // empty')
+
+  # Protected paths
+  PROTECTED_PATHS=(
+    "^/etc/"
+    "^/usr/"
+    "^/bin/"
+    "^/sbin/"
+    "^/boot/"
+    "^/sys/"
+    "^/proc/"
+    "\.env$"
+    "\.env\.local$"
+    "credentials"
+    "secrets"
+    "\.pem$"
+    "\.key$"
+    "id_rsa"
+    "id_ed25519"
+  )
+
+  for pattern in "${PROTECTED_PATHS[@]}"; do
+    if echo "$FILE_PATH" | grep -qiE "$pattern"; then
+      echo "[PreToolUse] BLOCKED: Protected file path" >&2
+      echo "[PreToolUse] Path: $FILE_PATH" >&2
+      echo '{"decision": "block", "reason": "Protected file path blocked by security hook"}'
+      exit 0
+    fi
+  done
+fi
+
+# Allow the tool to proceed
+echo '{"decision": "allow"}'
+exit 0

package/assets/.claude/settings.json

@@ -0,0 +1,28 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "permissions": {
+    "allow": [
+      "Bash(python -m pytest:*)",
+      "Bash(make:*)",
+      "Bash(git status)",
+      "Bash(git diff:*)",
+      "Bash(git log:*)"
+    ],
+    "deny": []
+  },
+  "hooks": {
+    "pre_tool_use": [
+      {
+        "matcher": "Bash",
+        "command": "python hooks/pre_tool_use.py"
+      }
+    ],
+    "post_tool_use": [
+      {
+        "matcher": "Bash",
+        "command": "python hooks/post_tool_use.py"
+      }
+    ]
+  },
+  "env": {}
+}

package/assets/.claude/settings.local.json

@@ -0,0 +1,7 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(dir:*)"
+    ]
+  }
+}

package/assets/.claude/skills/code-review.md

@@ -0,0 +1,20 @@
+---
+name: code-review
+description: Review code for quality, security, and best practices
+path: /review
+---
+
+# Code Review Skill
+
+You are a code reviewer. When the user asks you to review code, follow these steps:
+
+1. **Read the code** carefully
+2. **Check for issues**:
+   - Security vulnerabilities (injection, XSS, etc.)
+   - Performance problems
+   - Code style and readability
+   - Logic errors
+3. **Provide feedback** with specific line references
+4. **Suggest improvements** with code examples
+
+Be concise and actionable.

package/assets/.claude/skills/git-commit.md

@@ -0,0 +1,22 @@
+---
+name: git-commit
+description: Create well-formatted git commits with conventional commit messages
+path: /commit
+---
+
+# Git Commit Skill
+
+When the user asks you to commit changes:
+
+1. Run `git status` to see changes
+2. Run `git diff` to understand what changed
+3. Stage appropriate files with `git add`
+4. Create a commit message following conventional commits:
+   - `feat:` for new features
+   - `fix:` for bug fixes
+   - `docs:` for documentation
+   - `refactor:` for code refactoring
+   - `test:` for tests
+   - `chore:` for maintenance
+
+Keep the subject line under 50 characters.

package/assets/.claude/skills/harness-maintenance/SKILL.md

@@ -0,0 +1,45 @@
+---
+name: harness-maintenance
+description: Maintain the long-run harness files (harness/*). Use at session start/end and after meaningful milestones.
+---
+
+# harness-maintenance
+
+## Files to maintain
+
+- harness/PROGRESS.md
+- harness/FEATURES.json
+- harness/WORKLOG.md
+- harness/DECISIONS.md
+- harness/RISKS.md
+
+## Update rules
+
+### PROGRESS.md
+
+- Keep: Today Goal, Current Status, Next Actions (max 5), Blockers.
+- Update at session start and session end.
+
+### FEATURES.json
+
+- Keep features small and structured.
+- Each feature should include at least:
+  - id, title, status (todo|doing|done|blocked)
+  - **spec_file**: path to source spec file (e.g., "spec/01-auth.md")
+  - **priority**: numeric priority from spec file
+  - acceptance_criteria (array of strings)
+  - tasks (array of strings)
+  - tests (array of strings)
+  - **depends_on**: array of feature IDs (optional)
+- Update when scope changes, tasks are completed, or new work is discovered.
+- When a new spec file is added, create a corresponding feature entry.
+
+### WORKLOG.md
+
+- Append short bullets per day.
+- Record: what changed, what was fixed, tests run, important outcomes.
+
+### DECISIONS.md / RISKS.md
+
+- Only add when it's meaningful (avoid noise).
+- Keep each entry short and actionable.

package/assets/.claude/skills/quality-gates/SKILL.md

@@ -0,0 +1,17 @@
+---
+name: quality-gates
+description: Run quick checks and full CI tests. Use after edits, before finishing, or when tests fail.
+---
+
+# quality-gates
+
+## Commands
+
+- Quick: ./scripts/quickcheck.sh
+- Full : ./scripts/ci.sh
+
+## Rules
+
+- If tests fail: fix and re-run until green.
+- Do not claim completion without green full test run.
+- Prefer minimal fixes aligned with acceptance criteria.

package/assets/.claude/skills/quality-gates/scripts/stop_gate.sh

@@ -0,0 +1,69 @@
+#!/usr/bin/env bash
+# Stop Hook: Quality Gate
+# Runs before Claude can declare "done"
+# Ensures: tests pass, lint passes, no security issues
+
+set -euo pipefail
+
+cd "${CLAUDE_PROJECT_DIR:-.}"
+
+echo "========================================"
+echo "[stop_gate] Quality Gate Check"
+echo "========================================"
+
+FAILED=0
+
+# === 1. LINT / TYPE CHECK ===
+if [[ -x "./scripts/lint.sh" ]]; then
+  echo ""
+  echo "[stop_gate] Step 1/3: Lint & Type Check"
+  echo "----------------------------------------"
+  if ! ./scripts/lint.sh; then
+    echo "[stop_gate] ERROR: Lint/type check failed" >&2
+    FAILED=1
+  fi
+else
+  echo "[stop_gate] Step 1/3: Lint (skipped - no scripts/lint.sh)"
+fi
+
+# === 2. SECURITY SCAN ===
+if [[ -x "./scripts/security_scan.sh" ]]; then
+  echo ""
+  echo "[stop_gate] Step 2/3: Security Scan"
+  echo "----------------------------------------"
+  if ! ./scripts/security_scan.sh; then
+    echo "[stop_gate] WARNING: Security scan found issues" >&2
+    # Security issues are warnings, not blockers (can be false positives)
+    # Change to FAILED=1 if you want to block on security issues
+  fi
+else
+  echo "[stop_gate] Step 2/3: Security (skipped - no scripts/security_scan.sh)"
+fi
+
+# === 3. TESTS ===
+if [[ -x "./scripts/ci.sh" ]]; then
+  echo ""
+  echo "[stop_gate] Step 3/3: Full Test Suite"
+  echo "----------------------------------------"
+  if ! ./scripts/ci.sh; then
+    echo "[stop_gate] ERROR: Tests failed" >&2
+    FAILED=1
+  fi
+else
+  echo "[stop_gate] Step 3/3: Tests (skipped - no scripts/ci.sh)"
+  echo "[stop_gate] WARNING: No test script found. Create scripts/ci.sh" >&2
+fi
+
+# === SUMMARY ===
+echo ""
+echo "========================================"
+if [[ $FAILED -eq 0 ]]; then
+  echo "[stop_gate] All quality gates PASSED"
+  echo "========================================"
+  exit 0
+else
+  echo "[stop_gate] Quality gates FAILED"
+  echo "========================================"
+  echo "[stop_gate] Fix the issues above and try again." >&2
+  exit 2
+fi

package/assets/.claude/skills/req-to-plan/SKILL.md

@@ -0,0 +1,19 @@
+---
+name: req-to-plan
+description: Convert spec/REQ.md into a concrete implementation plan. Use when a requirements doc is provided or scope needs breakdown.
+---
+
+# req-to-plan
+
+## Output
+
+- docs/IMPLEMENTATION_PLAN.md
+
+## Must include
+
+- Goal / Non-goals
+- Acceptance Criteria (testable checklist)
+- Ordered task list
+- Impacted files/modules (likely)
+- Test plan (unit/integration/e2e)
+- Risks & assumptions

package/assets/.claude/skills/review-checklist/SKILL.md

@@ -0,0 +1,27 @@
+---
+name: review-checklist
+description: Review checklist for PR-style diffs. Use after code changes. Focus on correctness, security, maintainability, and tests.
+---
+
+# review-checklist
+
+## P0 / Must fix
+
+- Incorrect behavior vs acceptance criteria
+- Security issues (authz, injection, secrets, PII logging)
+- Breaking changes without migration/backward compatibility
+- Missing tests for changed logic
+- Flaky tests / nondeterminism introduced
+
+## P1 / Should fix
+
+- Poor error handling / missing validation
+- Hard-to-maintain structure (high complexity)
+- Performance regressions in hot paths
+- Inconsistent naming / unclear responsibilities
+
+## Output format
+
+- Summary
+- Findings (P0/P1/P2)
+- Test gaps

package/assets/.claude/skills/test-strategy/SKILL.md

@@ -0,0 +1,16 @@
+---
+name: test-strategy
+description: Define test strategy based on acceptance criteria. Use when planning changes or when test coverage is unclear.
+---
+
+# test-strategy
+
+## Guidance
+
+- Map each Acceptance Criteria to at least one test (unit/integration/e2e).
+- Prefer unit tests for pure logic; integration tests for IO/DB/API boundaries; e2e only for critical flows.
+- Include negative and edge cases (validation, authz, empty inputs, large inputs).
+
+## Output suggestions
+
+- Add a "Test Plan" section in docs/IMPLEMENTATION_PLAN.md.

package/dist/commands/init.js

@@ -0,0 +1,59 @@
+import fs from "fs";
+import path from "path";
+import { fileURLToPath } from "url";
+import chalk from "chalk";
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const ASSETS_DIR = path.resolve(__dirname, "../../assets");
+export async function init(options) {
+    const targetDir = process.cwd();
+    const claudeDir = path.join(targetDir, ".claude");
+    console.log(chalk.blue("Claude Code Bootstrap"));
+    console.log();
+    // Check if .claude already exists
+    if (fs.existsSync(claudeDir) && !options.force) {
+        console.log(chalk.yellow("⚠ .claude directory already exists."));
+        console.log(chalk.gray("  Use --force to overwrite."));
+        process.exit(1);
+    }
+    // Copy assets
+    const sourceDir = path.join(ASSETS_DIR, ".claude");
+    if (!fs.existsSync(sourceDir)) {
+        console.log(chalk.red("✗ Assets not found. Please reinstall the package."));
+        process.exit(1);
+    }
+    console.log(chalk.gray("Installing to: " + targetDir));
+    console.log();
+    try {
+        copyDirSync(sourceDir, claudeDir);
+        console.log(chalk.green("✓ .claude/settings.json"));
+        console.log(chalk.green("✓ .claude/skills/"));
+        console.log(chalk.green("✓ .claude/agents/"));
+        console.log(chalk.green("✓ .claude/commands/"));
+        console.log(chalk.green("✓ .claude/hooks/"));
+        console.log();
+        console.log(chalk.green("✓ Claude Code Bootstrap installed successfully!"));
+        console.log();
+        console.log(chalk.gray("Next steps:"));
+        console.log(chalk.gray("  1. Open the project in VS Code / Cursor"));
+        console.log(chalk.gray("  2. Start Claude Code"));
+        console.log(chalk.gray("  3. Use /start-session to begin"));
+    }
+    catch (error) {
+        console.log(chalk.red("✗ Failed to install: " + error.message));
+        process.exit(1);
+    }
+}
+function copyDirSync(src, dest) {
+    fs.mkdirSync(dest, { recursive: true });
+    const entries = fs.readdirSync(src, { withFileTypes: true });
+    for (const entry of entries) {
+        const srcPath = path.join(src, entry.name);
+        const destPath = path.join(dest, entry.name);
+        if (entry.isDirectory()) {
+            copyDirSync(srcPath, destPath);
+        }
+        else {
+            fs.copyFileSync(srcPath, destPath);
+        }
+    }
+}

package/dist/index.js

@@ -0,0 +1,13 @@
+#!/usr/bin/env node
+import { Command } from "commander";
+import { init } from "./commands/init.js";
+const program = new Command();
+program
+    .name("create-ccb")
+    .description("Claude Code Bootstrap - Setup your Claude Code configuration")
+    .version("1.0.0");
+program
+    .option("-f, --force", "Overwrite existing files")
+    .option("-v, --ver <version>", "Install specific version")
+    .action(init);
+program.parse();

package/package.json

@@ -0,0 +1,33 @@
+{
+  "name": "create-ccb",
+  "version": "1.0.0",
+  "description": "CLI to install Claude Code Bootstrap configuration",
+  "type": "module",
+  "bin": {
+    "create-ccb": "./dist/index.js"
+  },
+  "scripts": {
+    "build": "bun build src/index.ts --outdir dist --target node",
+    "dev": "bun run src/index.ts"
+  },
+  "files": [
+    "dist",
+    "assets"
+  ],
+  "keywords": [
+    "claude",
+    "claude-code",
+    "bootstrap",
+    "cli"
+  ],
+  "author": "ryoryoai",
+  "license": "MIT",
+  "dependencies": {
+    "commander": "^12.0.0",
+    "chalk": "^5.3.0"
+  },
+  "devDependencies": {
+    "@types/node": "^20.0.0",
+    "typescript": "^5.0.0"
+  }
+}