Claude Worktree Manager

Claude Worktree Manager

Automated worktree management for Claude Code development sessions with smart naming, auto-setup, and cleanup.

How to Use This Skill

There are two ways to create worktrees:

1. Via Claude Code Skill (Recommended)

Use the Skill tool to invoke this skill - the script path is detected automatically:

// Using Skill tool with args parameter (Recommended)
Skill(skill: 'claude-worktree-manager', args: 'create feature-name --model opus')
Skill(skill: 'claude-worktree-manager', args: 'create feature-name --isolated --model sonnet')

// All bash script flags work the same via args
Skill(skill: 'claude-worktree-manager', args: 'list')
Skill(skill: 'claude-worktree-manager', args: 'cleanup --days 14')

The skill handles all path resolution using git rev-parse --show-toplevel, so you only need to provide the worktree name and optional flags. All bash script flags (--model, --isolated, --days) work exactly the same when passed via the args parameter.

2. Via Direct Script Invocation

If you're running the script manually, provide the full path from any directory:

.claude/skills/claude-worktree-manager/scripts/worktree.sh create feature-name --model opus

Or use the full absolute path:

/path/to/repo/.claude/skills/claude-worktree-manager/scripts/worktree.sh create feature-name --model opus

Invocation Method Comparison

To clarify the correct syntax for each invocation method:

Key Differences:

• Skill tool: All commands and flags are passed as a single string via the args parameter
• Direct script: Commands and flags are bash arguments, requires absolute or relative path

Discovery-First Workflow (Recommended)

For non-trivial features, use the discovery-first workflow with the /discover skill before creating a worktree:

Workflow Overview

┌─────────────────────────────────────────────────────────────┐
│  MAIN BRANCH                                                │
│                                                             │
│  1. /discover <feature-description>                         │
│     ├── Agent asks clarifying questions (one at a time)     │
│     ├── Proposes 2-3 approaches                             │
│     ├── Validates design in sections                        │
│     └── Creates docs/plans/YYYY-MM-DD-<feature>.md          │
│                                                             │
│  2. Agent offers to create worktree with --plan             │
│                                                             │
└─────────────────────────────────────────────────────────────┘
                          │
                          ▼
┌─────────────────────────────────────────────────────────────┐
│  WORKTREE (created with --plan)                             │
│                                                             │
│  3. Worktree agent reads plan and executes tasks            │
│     ├── Two-stage review (spec compliance + code quality)   │
│     ├── Mandatory verification before completion claims     │
│     └── Batch checkpoints for human feedback                │
│                                                             │
└─────────────────────────────────────────────────────────────┘

Example: Discovery to Worktree

# Step 1: Run discovery in main branch
User: /discover add webhook support for integrations

Agent: [Explores codebase]
Agent: [AskUserQuestion - one question at a time]
       "What events should trigger webhooks?"
       □ Integration lifecycle (connected, disconnected)
       □ Agent actions (tool calls, completions)
       □ All events

User: "Integration lifecycle"

Agent: [More questions, design validation...]
Agent: [Writes plan to docs/plans/2026-02-01-webhooks.md]
Agent: "Plan complete. Ready to create worktree?"

User: "Yes"

# Step 2: Create worktree with plan
Agent: Skill(skill: 'claude-worktree-manager',
             args: 'create webhooks --plan docs/plans/2026-02-01-webhooks.md')

The --plan Flag

The --plan flag accepts a path to an implementation plan file:

--plan docs/plans/2026-02-01-feature-name.md

What it does:

1. Verifies plan exists - Checks if the plan file exists in the repo
2. Auto-constructs goal - Creates a goal that instructs the worktree agent to:
   • Read the plan first
   • Execute tasks in order
   • Use parallel subagents for 4+ independent tasks
   • Apply two-stage review (spec compliance, then code quality)
   • Provide verification evidence for every completion claim

Auto-generated goal format:

Implement the plan at docs/plans/2026-02-01-webhooks.md

Read the plan first, then execute tasks in order. For 4+ independent tasks,
consider using parallel subagents. Use two-stage review (spec compliance then
code quality). Mandatory: every completion claim must include verification output.

Combining --plan with Other Flags

# Plan + specific model for complex work
Skill(skill: 'claude-worktree-manager',
      args: 'create webhooks --plan docs/plans/2026-02-01-webhooks.md --model opus')

# Plan + isolated database for schema changes
Skill(skill: 'claude-worktree-manager',
      args: 'create migrations --plan docs/plans/2026-02-01-db-changes.md --isolated')

# Plan + goal (additional context appended)
Skill(skill: 'claude-worktree-manager',
      args: 'create webhooks --plan docs/plans/2026-02-01-webhooks.md --goal "Focus on error handling first"')

When to Use --plan vs --goal

Related Skills

• /discover - Creates the plan through structured dialogue
• plan-executor - Guides worktree agent on executing plans with verification

Verifying Model Configuration

After creating a worktree with --model, verify the configuration was set correctly:

# Navigate to your worktree
cd /path/to/worktree

# Run verification script
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh

# Or verify from anywhere by passing the path
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh /path/to/worktree

Expected output when successful:

[INFO] Verifying Claude model configuration in: /path/to/worktree
[SUCCESS] .claude/settings.local.json found
[SUCCESS] Model configuration found: opus
[SUCCESS] Model is valid: opus

The validation script will:

1. Check that .claude/settings.local.json exists
2. Extract the model value using jq or grep
3. Validate the model is one of: opus, sonnet, haiku
4. Provide instructions if configuration is missing or invalid

See the Model Flag Configuration - Edge Cases & Troubleshooting section below for fixing common issues.

⚠️ Critical: Model Configuration Warning

Before starting Claude in a worktree with --model flag, you MUST run the verification script.

The model configuration uses jq (preferred) or sed (fallback) to update JSON files. These tools can fail silently on some systems, resulting in Claude using the wrong model.

Quick Reference: Model Configuration Troubleshooting

Default Model Configuration

The script has a default model configured: sonnet

• All new worktrees automatically use this model unless overridden with --model
• To change the default, edit DEFAULT_MODEL in the script
• Override for specific worktrees: --model opus or --model haiku

⚠️ Important: Setting DEFAULT_MODEL does NOT make verification optional!

Even with a default model configured, you MUST still run verify-worktree-model.sh after creating a worktree. The default model goes through the same jq/sed configuration process that can fail silently.

Complete Workflow: Create → Verify → Start

Scenario 1: Using default model (sonnet)

# 1. Create worktree (automatically uses DEFAULT_MODEL=sonnet)
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature | tail -1)
# Output: [INFO] Using default model: sonnet

# 2. REQUIRED: Verify the default model was set correctly
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh "$WORKTREE_PATH"
# Expected: [SUCCESS] Model is valid: sonnet

# 3. Only start Claude AFTER verification succeeds
cd "$WORKTREE_PATH" && claude 'Your goal'

Scenario 2: Using custom model (opus)

# 1. Create worktree with explicit model (overrides DEFAULT_MODEL)
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create complex-task --model opus | tail -1)
# Output: [INFO] Setting default model to: opus

# 2. REQUIRED: Verify the custom model was set correctly
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh "$WORKTREE_PATH"
# Expected: [SUCCESS] Model is valid: opus

# 3. Only start Claude AFTER verification succeeds
cd "$WORKTREE_PATH" && claude 'Your complex goal'

Verification output for default model worktrees:

[INFO] Verifying Claude model configuration in: /path/to/worktree
[SUCCESS] .claude/settings.local.json found
[SUCCESS] Model configuration found: sonnet
[SUCCESS] Model is valid: sonnet

Quick Start

Create a Worktree

When the user asks to create a worktree for a feature or environment, derive a short, descriptive kebab-case name from their request:

Naming examples:

• "staging environment" → staging-env
• "add dark mode toggle" → dark-mode
• "fix authentication bug" → fix-auth-bug
• "update API endpoints" → update-api

Then run:

# Standard (uses shared dev database)
.claude/skills/claude-worktree-manager/scripts/worktree.sh create <derived-name>

# With specific model (e.g., opus, sonnet, haiku)
.claude/skills/claude-worktree-manager/scripts/worktree.sh create <derived-name> --model opus

# Isolated (creates dedicated database with seeding)
.claude/skills/claude-worktree-manager/scripts/worktree.sh create <derived-name> --isolated

# Isolated with model
.claude/skills/claude-worktree-manager/scripts/worktree.sh create <derived-name> --isolated --model sonnet

What happens:

1. Cleans up worktrees older than 7 days
2. Fetches latest from origin
3. Checks if a branch with your name already exists on origin:
   • If found: Creates local branch tracking the remote, pulls latest changes
   • If not found: Creates new branch: worktree/<name>-<timestamp> from main
4. Creates worktree at: ~/claude-worktrees/<project-name>/<name>-<timestamp>
5. Copies .env from main repo
6. Copies .claude/settings.local.json from main repo (API keys, preferences)
7. Starts pnpm install in background
8. If --model: Sets default Claude model in .claude/settings.local.json
9. If --isolated: Creates dedicated database and seeds it with test data
10. Returns the worktree path immediately

Required: Verify Model Configuration Before Starting Claude

When creating worktrees with the --model flag, you MUST verify the configuration before starting Claude:

# 1. Create worktree with model
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature --model opus | tail -1)

# 2. REQUIRED: Verify the model was set correctly
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh "$WORKTREE_PATH"
# Must show: [SUCCESS] Model is valid: opus

# 3. Check pnpm install progress (optional)
tail -f "$WORKTREE_PATH/.pnpm-install.log"

# 4. Only start Claude AFTER verification succeeds
cd "$WORKTREE_PATH"
claude 'Your goal here'

Why is verification required?

1. jq/sed fallback can fail silently: The script uses jq (preferred) or sed (fallback) to update JSON. On some systems, sed may fail without error.

2. macOS vs Linux sed differences: macOS uses BSD sed which requires -i '' while Linux uses GNU sed with -i. The script handles this, but edge cases exist.

3. JSON syntax can break: Malformed JSON in settings.local.json can cause the update to fail.

4. Claude uses wrong model without verification: If you start Claude before verifying, it will use the default model instead of your requested model, wasting time and potentially money.

If verification fails: See the "Model Flag Configuration - Edge Cases & Troubleshooting" section below for manual fixes.

When to Use --isolated

Use the --isolated flag when:

• Schema changes: Testing database migrations
• Migration testing: Verifying migration scripts work correctly
• Isolated experiments: Need a clean database state
• Breaking changes: Don't want to affect shared dev data

For normal feature development, skip --isolated to use the shared database.

Automatic Script Location Detection

When you invoke this skill using the Skill tool (e.g., Skill(claude-worktree-manager)), the skill automatically:

1. Detects your current git repository using git rev-parse --show-toplevel
2. Locates the worktree script at .claude/skills/claude-worktree-manager/scripts/worktree.sh
3. Executes the script from the repository root

You don't need to specify absolute paths or worry about the script location. The skill handles all path resolution automatically. Just use simple worktree names and optional flags:

# Via Skill tool - automatic path resolution
Skill(claude-worktree-manager) create my-feature --model sonnet

# Via direct script - requires full path from any directory
/path/to/repo/.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature --model sonnet

Model Selection

Use the --model flag to set the default Claude model for the worktree:

.claude/skills/claude-worktree-manager/scripts/worktree.sh create <name> --model <model-name>

Available models: opus, sonnet, haiku

When specified, the model is configured in the worktree's .claude/settings.local.json so that all Claude Code sessions in that worktree use the selected model by default. This is useful for:

• Testing with specific models: Quickly test a feature with Opus for complex work or Haiku for speed
• Cost optimization: Use Haiku for routine tasks, Opus for complex reasoning
• Consistency: Ensure all team members working in a specific worktree use the same model

Important: Return the worktree path to the user so they can open a new Claude Code session in that directory.

Ghostty Integration (Automatic)

The script automatically opens a new Ghostty tab when running in Ghostty terminal. Use the --goal flag to set a task description:

# Creates worktree and opens Ghostty tab with Claude
.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature --goal "Add dark mode toggle"

The script will:

1. Create the worktree with all setup
2. Detect if running in Ghostty ($TERM_PROGRAM = ghostty)
3. Open a new tab using ghostty-tab
4. Navigate to the worktree directory
5. Type the Claude command (without executing, so you can review)

Goal Prompt Guidelines:

• Keep it concise (1-2 sentences max)
• Focus on the task objective
• Use imperative voice
• Avoid special characters that need escaping (use simple quotes)

Examples:

.claude/skills/claude-worktree-manager/scripts/worktree.sh create fix-auth --goal "Fix authentication redirect bug"
.claude/skills/claude-worktree-manager/scripts/worktree.sh create oauth --goal "Add OAuth2 support with Google and GitHub"
.claude/skills/claude-worktree-manager/scripts/worktree.sh create refactor --goal "Refactor to multi-tenant architecture"

Script Configuration

The script has configurable variables at the top. To customize, edit the script directly:

# Edit the script
nano .claude/skills/claude-worktree-manager/scripts/worktree.sh

Configuration variables:

Recommended: Configure your shell alias instead

Rather than editing script variables, configure your Claude alias in ~/.zshrc:

# Claude Code alias with default model
alias cc="claude --dangerously-skip-permissions --model sonnet"

This way, all Claude sessions (not just worktrees) use your preferred defaults.

Plan Mode: When you provide a --goal, the script automatically wraps it as:

enter plan mode to work on: <your goal>

This instructs Claude to enter plan mode and begin planning the task.

Important: The ghostty-tab command is at ~/.local/bin/ghostty-tab and uses AppleScript to open a new Ghostty tab. It accepts:

• -d <path> - Directory to cd into
• --no-enter - Type the command but don't press enter (lets user review before executing)
• "<command>" - Command to type/run after cd

List Worktrees

.claude/skills/claude-worktree-manager/scripts/worktree.sh list

Shows all active worktrees for the current project with their branches.

Manual Cleanup

# Cleanup worktrees older than 7 days (default)
.claude/skills/claude-worktree-manager/scripts/worktree.sh cleanup

# Custom age threshold
.claude/skills/claude-worktree-manager/scripts/worktree.sh cleanup --days 14

Origin Branch Handling

The script intelligently checks if a branch with your name already exists on origin:

Scenario 1: Branch exists on origin

.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature

If origin/my-feature exists, the script will:

• Create a local branch tracking origin/my-feature
• Pull the latest changes automatically
• Create the worktree on this branch
• Useful for continuing work on an existing feature branch

Scenario 2: Branch doesn't exist on origin

.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature

If origin/my-feature doesn't exist, the script will:

• Create a new branch from main: worktree/my-feature-<timestamp>
• Create a fresh worktree for new development

This means you can use simple names (like restructure-prompts) and the script will automatically handle both new work and continuing existing branches.

Workflow

1. User Requests Worktree

User says: "I want to create a new worktree for developing a staging environment"

2. Derive Smart Name

Analyze the request and derive a concise kebab-case name:

• Extract key purpose: "staging environment"
• Convert to kebab-case: staging-env

3. Create Worktree and Capture Path

# Run the script and capture the worktree path (last line of output)
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create staging-env | tail -1)
echo "Created: $WORKTREE_PATH"

4. Open in Ghostty (if available)

After creating the worktree, check if running in Ghostty and automatically open a new tab:

# Check if in Ghostty and open new tab with Claude (--no-enter lets user review first)
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Set up and configure the staging environment'"
fi

5. Return Path

The script outputs the worktree path. Return it to the user:

Worktree created at: ~/claude-worktrees/orienter/staging-env-1736639420

✓ Opened new Ghostty tab with Claude command ready (press Enter to start)
  Goal: Set up and configure the staging environment

Note: pnpm install is running in the background. Check progress with:
tail -f ~/claude-worktrees/orienter/staging-env-1736639420/.pnpm-install.log

If NOT in Ghostty, show manual instructions:

Worktree created at: ~/claude-worktrees/orienter/staging-env-1736639420

You can now open a new Claude Code session in this directory:
cd ~/claude-worktrees/orienter/staging-env-1736639420

Note: pnpm install is running in the background. Check progress with:
tail -f ~/claude-worktrees/orienter/staging-env-1736639420/.pnpm-install.log

Background Installation

The worktree creation starts pnpm install in the background using nohup. This means:

• The path is returned immediately (don't wait for pnpm)
• Installation runs async and logs to .pnpm-install.log
• User can start working right away
• Dependencies will be available after a few moments

Check if installation is complete:

# Check if still running
ps aux | grep pnpm

# Watch the log
tail -f <worktree-path>/.pnpm-install.log

Database Seeding (Isolated Mode)

When using --isolated, the script automatically:

1. Creates a new SQLite database file: data/worktree_<timestamp>.db
2. Updates the worktree's .env with the new SQLITE_DB_PATH
3. Runs all migrations
4. Seeds the database with:
   • 5 agents: pm-assistant, communicator, scheduler, explorer, app-builder
   • 6 context rules: Platform and environment routing
   • 6 test permissions: Sample permissions
   • 4 sample prompts: Default prompts for testing

Database seeding starts after pnpm install completes. Check progress:

tail -f <worktree-path>/.db-seed.log

Manual Database Seeding

If you need to seed an existing worktree:

# Seed with shared database (from .env)
./scripts/seed-worktree-db.sh

# Create isolated database and seed
ISOLATED=true ./scripts/seed-worktree-db.sh

Configuration Files

The script automatically handles configuration files:

Automatically Available (via git):

• .claude/skills/ - All skills are available in worktrees
• .claude/settings.json - Committed Claude settings

Automatically Copied:

• .env - Environment variables for the application
• .claude/settings.local.json - Local Claude settings (API keys, preferences)

This ensures your worktree has the same development environment as the main repo.

Integration with worktree-operations

After creating the worktree, users can reference the worktree-operations skill for:

• Building packages: pnpm run build
• Running tests: pnpm test
• Development mode: pnpm run dev
• Type checking: pnpm run typecheck

The worktree-operations skill covers all pnpm/turbo commands for working in the monorepo.

Directory Structure

~/claude-worktrees/
└── <project-name>/
    ├── feature-a-1736639420/
    ├── staging-env-1736639421/
    └── fix-bug-1736639422/

Each project gets its own subdirectory under ~/claude-worktrees/.

Cleanup Policy

• Automatic: Runs before each worktree creation
• Threshold: Removes worktrees older than 7 days (based on modification time)
• Safe: Only removes worktrees in the Claude worktree directory
• Manual: Can be triggered anytime with ./worktree.sh cleanup

.env Configuration Requirements

The .env file is copied to worktrees automatically. Ensure proper formatting to avoid shell parsing errors:

Quote Special Characters

Values with special characters MUST be quoted:

# ✅ Correct - quoted values
SQLITE_DB_PATH="./data/orient.db"
STANDUP_CRON="30 9 * * 1-5"
STANDUP_CHANNEL="#orienter-standups"

# ❌ Wrong - unquoted special chars cause shell errors
STANDUP_CRON=30 9 * * 1-5          # Shell expands * as glob
STANDUP_CHANNEL=#orienter-standups  # Shell treats # as comment

Required Variables

For database seeding to work, ensure SQLITE_DB_PATH is set:

SQLITE_DB_PATH="./data/orient.db"

Common Shell Parsing Errors

If you see errors like command not found when sourcing .env:

• Check for unquoted cron expressions (* * *)
• Check for unquoted channel names starting with #
• Ensure no trailing spaces after values

Model Flag Configuration - Edge Cases & Troubleshooting

The --model flag updates .claude/settings.local.json to set the default Claude model for your worktree. This section documents potential edge cases and how to handle them.

Requirements

The script requires one of the following tools to update JSON:

1. jq (preferred) - Safely parses and updates JSON
2. sed (fallback) - Uses text substitution, less reliable but works when jq is unavailable

Most systems have sed built-in. To install jq:

# macOS
brew install jq

# Ubuntu/Debian
sudo apt-get install jq

# Or use your system package manager

How Model Configuration Works

The script attempts to update .claude/settings.local.json with your selected model:

Step 1: Check for jq

if command -v jq &> /dev/null; then
    # Use jq (preferred, safe JSON manipulation)
else
    # Fall back to sed (text substitution, can fail silently)
fi

Step 2a: With jq (preferred)

jq ".model = \"opus\"" settings.local.json > settings.local.json.tmp
mv settings.local.json.tmp settings.local.json

• ✅ Safely parses JSON
• ✅ Preserves formatting and other keys
• ✅ Creates valid JSON output
• ✅ Handles edge cases (nested objects, arrays)

Step 2b: With sed (fallback)

# macOS (BSD sed)
sed -i '' "s/\"model\": \"[^\"]*\"/\"model\": \"opus\"/g" file.json

# Linux (GNU sed)
sed -i "s/\"model\": \"[^\"]*\"/\"model\": \"opus\"/g" file.json

• ⚠️ Text-based replacement, not JSON-aware
• ⚠️ May fail silently if pattern doesn't match
• ⚠️ macOS and Linux have different sed flags

macOS vs Linux sed Differences

The script handles these differences automatically, but edge cases can still occur.

JSON File States: Valid vs Broken

✅ Valid JSON after successful configuration:

{
  "model": "opus",
  "permissions": {
    "allow": ["Bash(git:*)"]
  }
}

❌ Broken JSON - sed added model incorrectly:

{{"model": "opus",
  "permissions": {
    "allow": ["Bash(git:*)"]
  }
}

❌ Broken JSON - duplicate model keys:

{
  "model": "opus",
  "model": "sonnet",
  "permissions": { ... }
}

❌ Missing model - sed pattern didn't match:

{
  "permissions": {
    "allow": ["Bash(git:*)"]
  }
}

Troubleshooting: Model Not Applied

If Claude doesn't use the configured model:

Step 1: Verify the model was set

# Go to your worktree and verify
cd /path/to/worktree
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh

# Or from anywhere (pass worktree path)
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh /path/to/worktree

Step 2: Check settings.local.json contents

cat /path/to/worktree/.claude/settings.local.json | grep -i model

You should see:

{
  "model": "opus",
  "permissions": { ... }
}

Step 3: If model is missing, add it manually

If the model key is missing, add it to the top level of .claude/settings.local.json:

# Edit the file manually
nano /path/to/worktree/.claude/settings.local.json

The file should look like:

{
  "model": "opus",
  "permissions": {
    "allow": [ ... ]
  }
}

Common Issues

Issue: "jq not found" in logs but model was set

Why: The script fell back to sed, which is less reliable. jq is recommended.

Fix: Install jq and re-create the worktree with the --model flag

# Install jq
brew install jq  # or sudo apt-get install jq

# Re-create worktree with model flag
.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature --model opus

Issue: Model key added but not using that model

Why: The model might be formatted incorrectly or Claude Code might not be reading it.

Fix: Verify the JSON is valid and properly formatted:

# Verify JSON syntax is valid
jq . /path/to/worktree/.claude/settings.local.json

# Should output valid JSON without errors

If JSON is invalid, edit the file to fix syntax errors (missing quotes, commas, etc).

Issue: sed failed silently, model not set

Why: This can happen on systems with different sed versions or in edge cases.

Fix: Use jq or manually edit the file

# Option 1: Install jq and re-create worktree
brew install jq
.claude/skills/claude-worktree-manager/scripts/worktree.sh create my-feature --model opus

# Option 2: Manually add the model to the file
nano /path/to/worktree/.claude/settings.local.json
# Add "model": "opus", at the top level

Issue: Model works in worktree but not when using goal prompt

Why: The Ghostty tab might be starting Claude in a different context or with different settings.

Fix: Verify the model is set and start Claude from within the worktree:

# Navigate to worktree
cd /path/to/worktree

# Verify model
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh

# Start Claude directly
claude 'Your goal prompt here'

# Claude will load settings from the current worktree directory

Validation Script

Use the validation script to verify worktree model configuration:

# From the worktree directory
cd /path/to/worktree
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh

# Or from anywhere
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh /path/to/worktree

Output examples:

✅ Correct:

[INFO] Verifying Claude model configuration in: /path/to/worktree
[SUCCESS] .claude/settings.local.json found
[SUCCESS] Model configuration found: opus
[SUCCESS] Model is valid: opus

❌ Missing model:

[INFO] Verifying Claude model configuration in: /path/to/worktree
[SUCCESS] .claude/settings.local.json found
[WARN] No model configuration found in settings.local.json

Manual Model Configuration

If automatic configuration fails, manually edit .claude/settings.local.json:

# Navigate to worktree
cd /path/to/worktree

# Open settings file
nano .claude/settings.local.json

Add the model key at the top level:

{
  "model": "opus",
  "permissions": {
    "allow": [ ... ]
  }
}

Valid model values:

• "opus" - Claude Opus 4.5 (most capable)
• "sonnet" - Claude Sonnet (balanced)
• "haiku" - Claude Haiku (fastest/cheapest)

After saving, verify with the validation script:

.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh

Troubleshooting

pnpm install failed

Check the log:

cat <worktree-path>/.pnpm-install.log

Re-run manually if needed:

cd <worktree-path>
pnpm install

.env not copied

Copy manually:

cp <main-repo>/.env <worktree-path>/.env

Database seeding failed

Check the log:

cat <worktree-path>/.db-seed.log

Common issues:

• SQLite database directory not writable: Check permissions on data/ directory
• Invalid SQLITE_DB_PATH: Check quotes and format
• Missing tables: Run npm run db:migrate first

Worktree creation failed

Check git status and ensure:

• You're in a git repository
• Remote is accessible
• No uncommitted changes blocking the operation

Examples

Create worktree for a feature (with Ghostty)

User: "Create a worktree for adding OAuth support"

# 1. Create the worktree and capture path
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create oauth-support | tail -1)

# 2. If in Ghostty, open new tab with Claude
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Add OAuth support with Google and GitHub providers'"
fi

Create worktree with specific model

User: "Create a worktree for complex refactoring using Opus"

# 1. Create the worktree with Opus model and capture path
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create complex-refactor --model opus | tail -1)

# 2. Verify the model was set correctly
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh "$WORKTREE_PATH"
# Expected output: [SUCCESS] Model is valid: opus

# 3. If in Ghostty, open new tab with Claude
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Refactor the project to support multi-tenant architecture'"
fi

User: "Create a worktree for quick bug fixes using Haiku"

# 1. Create the worktree with Haiku model
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create quick-fixes --model haiku | tail -1)

# 2. Verify the model was set correctly
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh "$WORKTREE_PATH"

# 3. Start working in the worktree
cd "$WORKTREE_PATH"
claude 'Fix critical bugs'

User: "Create an Opus worktree with isolated database for migration testing"

# 1. Create worktree with both flags
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create schema-testing --isolated --model opus | tail -1)

# 2. Verify model configuration (database seeding runs in parallel)
.claude/skills/claude-worktree-manager/scripts/verify-worktree-model.sh "$WORKTREE_PATH"
# Output should show: [SUCCESS] Model is valid: opus

# 3. Check database seeding progress (runs after pnpm install)
tail -f "$WORKTREE_PATH/.db-seed.log"

# 4. Once ready, navigate to worktree and start Claude
cd "$WORKTREE_PATH"
claude 'Test database migrations'

Create worktree for bug fix

User: "I need a worktree to fix the login redirect issue"

# 1. Create the worktree and capture path
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create fix-login-redirect | tail -1)

# 2. If in Ghostty, open new tab with Claude
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Fix the login redirect issue'"
fi

Create worktree for refactoring

User: "Create a new worktree for refactoring the project to make it multi-tenant"

# 1. Create the worktree and capture path
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create multi-tenant-refactor | tail -1)

# 2. If in Ghostty, open new tab with Claude
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Refactor the project to support multi-tenant architecture'"
fi

Create worktree with isolated database

User: "I need a worktree to test new database migrations"

# 1. Create the worktree with isolated DB and capture path
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create test-migrations --isolated | tail -1)

# 2. If in Ghostty, open new tab with Claude
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Test new database migrations in isolated environment'"
fi

Create worktree for schema changes

User: "Create a worktree for adding new agent tables"

# 1. Create the worktree with isolated DB and capture path
WORKTREE_PATH=$(.claude/skills/claude-worktree-manager/scripts/worktree.sh create add-agent-tables --isolated | tail -1)

# 2. If in Ghostty, open new tab with Claude
if [ "$TERM_PROGRAM" = "ghostty" ]; then
    ghostty-tab -d "$WORKTREE_PATH" --no-enter "claude 'Add new agent tables to the database schema'"
fi

List all worktrees

User: "Show me all my worktrees"

.claude/skills/claude-worktree-manager/scripts/worktree.sh list

Cleanup old worktrees

User: "Clean up worktrees older than 3 days"

.claude/skills/claude-worktree-manager/scripts/worktree.sh

…(truncated)