Nav Upgrade logo

Nav Upgrade

Organization
qf-studio
nav-upgrade

Automates Navigator plugin updates. Detects current version, updates plugin, verifies installation, updates project CLAUDE.md, and validates new features. Auto-invoke when user mentions upgrading Navigator or getting new features.

Overview

Publisherqf-studio
Repositorynavigator
Skill namenav-upgrade
Stars
232
Forks
12
Bundled files
6
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 6 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by qf-studio on GitHub. Read the source before you install it.

Installation

Install the Nav Upgrade AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/qf-studio/navigator.git /tmp/navigator
mkdir -p .claude/skills
cp -r /tmp/navigator/skills/nav-upgrade .claude/skills/nav-upgrade
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Nav Upgrade in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Nav Upgrade on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Nav Upgrade is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Navigator Upgrade Skill

Automate Navigator plugin updates with version detection, conflict resolution, and post-update validation.

When to Invoke

Auto-invoke when user says:

  • "Update Navigator"
  • "Upgrade Navigator plugin"
  • "Get latest Navigator version"
  • "Update to Navigator v3.3.0"
  • "Install new Navigator features"
  • "Check for Navigator updates"

What This Does

6-Step Workflow:

  1. Version Detection: Check current Navigator version vs latest
  2. Plugin Update: Execute /plugin update navigator
  3. Verification: Confirm update succeeded
  4. CLAUDE.md Update: Update project configuration (via nav-sync-claude)
  5. Hooks Setup: Install/update token monitoring hooks in project
  6. Feature Discovery: Show new features available

Time Savings: Manual update (10-15 min) → Automated (2 min)


Prerequisites

  • Navigator plugin installed
  • Project initialized with Navigator
  • Internet connection for plugin update

Workflow Protocol

Step 1: Version Detection

Execute: version_detector.py

Check both stable and pre-release versions:

bash
# Current installed version
grep '"version"' .claude-plugin/plugin.json

# Get all releases (including pre-releases)
curl -s https://api.github.com/repos/alekspetrov/navigator/releases

# Parse:
# - Latest stable (prerelease: false)
# - Latest pre-release (prerelease: true)
# - Compare with current version

Output scenarios:

Scenario 1: Stable update available

json
{
  "current_version": "4.0.0",
  "latest_stable": "4.2.0",
  "latest_prerelease": null,
  "recommendation": "update_to_stable"
}

Scenario 2: Pre-release available (user on stable)

json
{
  "current_version": "4.0.0",
  "latest_stable": "4.0.0",
  "latest_prerelease": "4.3.0",
  "recommendation": "offer_prerelease_option"
}

Present choice:

✅ You're on the latest stable version (v4.0.0)

⚡ Experimental version available: v4.3.0

New in v4.3.0 (Experimental):
• Multi-Claude agentic workflows
• 30% success rate (use for simple features)
• PM integration with ticket closing

Options:
[1] Stay on stable v4.0.0 (recommended)
[2] Try experimental v4.3.0 (early adopter)

Your choice [1-2]:

Scenario 3: Already on latest (stable or pre-release)

✅ You're on v4.3.0 (latest experimental)

Latest stable: v4.0.0
Status: You're ahead of stable (testing experimental features)

New features in your version:
- Multi-Claude workflows
- Task agents in sub-Claude phases

Skip to Step 5 (Feature Discovery).

Scenario 4: On pre-release, newer stable available

⚠️  You're on v4.3.0 (experimental)
Latest stable: v4.6.0

Recommendation: Update to stable v4.6.0
Experimental features from v4.3.0 are now stable.

Step 2: Plugin Update

Scenario-based update strategy:

Scenario 2: Pre-release Available (User on Stable)

When pre-release detected, present choice using AskUserQuestion tool:

markdown
✅ You're on latest stable version (v4.0.0)

⚡ Experimental version available: v4.3.0

New in v4.3.0 (Experimental):
• Multi-Claude agentic workflows
• 30% success rate (use for simple features)
• PM integration with ticket closing

**Question**: Which version would you like?

**Options**:
[1] **Stay on stable v4.0.0** (recommended)
    - Production-ready
    - No experimental features
    - Most reliable

[2] **Try experimental v4.3.0** (early adopter)
    - Multi-Claude workflows
    - Latest features
    - 30% completion rate
    - Help test new functionality

Your choice?

If user chooses [1] (Stay stable):

✓ Staying on v4.0.0 (latest stable)

No action needed. Run nav-upgrade again when you're ready to try experimental features.

If user chooses [2] (Try experimental):

bash
# Uninstall current version
/plugin uninstall navigator

# Add marketplace (if not already added)
/plugin marketplace add alekspetrov/navigator

# Install specific pre-release version
# Note: /plugin update only fetches stable, must install specific version
git clone https://github.com/alekspetrov/navigator.git /tmp/navigator-v4.3.0
cd /tmp/navigator-v4.3.0
git checkout v4.3.0

# Install from local checkout
/plugin install /tmp/navigator-v4.3.0

Then verify installation:

bash
/plugin list | grep navigator
# Should show: navigator (v4.3.0)
Scenario 1: Stable Update Available

Execute: /plugin update navigator

Monitor output:

Updating navigator...
✅ Navigator updated to v4.2.0

If update fails:

❌ Update failed: [error message]

Troubleshooting:
1. Restart Claude Code
2. Try: /plugin uninstall navigator && /plugin install navigator
3. Check internet connection
4. Report issue: https://github.com/alekspetrov/navigator/issues

Automatic retry (once): If update fails, try uninstall/reinstall automatically:

bash
/plugin uninstall navigator
/plugin marketplace add alekspetrov/navigator
/plugin install navigator

Step 3: Verification

Execute: plugin_verifier.py

Verify:

  1. Plugin version matches latest
  2. New skills registered in plugin.json
  3. Skills are invokable

Test new skills (v3.3.0 example):

bash
# Test that visual-regression skill exists
ls ~/.config/claude/plugins/navigator/skills/visual-regression/SKILL.md 2>/dev/null || echo "Skill not found"

Output:

✅ Update Verification

Version: v3.3.0 ✅
New Skills Registered: visual-regression ✅
Skills Invokable: ✅

Update successful!

If verification fails:

⚠️ Update completed but verification failed

Issue: visual-regression skill not found
Fix: Restart Claude Code to reload skills

After restarting, verify:
"Set up visual regression for Button"

Prompt user to restart Claude Code.

IMPORTANT - Restart Required After All Updates:

Claude Code caches skill paths at session start. Even successful updates require a restart to load new skills:

⚠️  RESTART REQUIRED

Plugin updated successfully to vX.Y.Z.
Claude Code caches skill paths at session start.

To use new/updated skills:
1. Close Claude Code
2. Reopen Claude Code
3. Run "nav start" to verify new version

Without restart:
- Old skill versions continue running
- New skills won't be available
- May see "skill not found" errors

This is a Claude Code behavior, not a Navigator issue. The plugin files update correctly, but the active session continues using cached skill paths from the old version.


Step 4: Update Project CLAUDE.md (Automatic)

After plugin update, automatically invoke: nav-sync-claude skill

🔄 Syncing project CLAUDE.md with updated plugin...

✓ Using template from GitHub (v4.3.0)
✓ Extracted customizations
✓ Generated updated CLAUDE.md

What happens automatically:

  1. Detects new plugin version (e.g., v4.3.0)
  2. Fetches matching template from GitHub
  3. Preserves project customizations
  4. Updates CLAUDE.md in current project
  5. Shows diff for review

Template sync benefits:

  • ✅ CLAUDE.md always matches installed plugin version
  • ✅ No template drift (v4.0 templates with v4.3 plugin)
  • ✅ Pre-release templates accessible
  • ✅ Offline fallback to bundled templates

User action required:

Review changes and commit:

git add CLAUDE.md
git commit -m "chore: update CLAUDE.md to Navigator v4.3.0"

See: nav-sync-claude skill for details.


Step 5: Migrate Legacy Hooks Out of .claude/settings.json

v6.13.0+: Navigator's lifecycle hooks now ship with the plugin manifest (.claude-plugin/plugin.json's top-level hooks field). Claude Code only substitutes ${CLAUDE_PLUGIN_ROOT} for hooks declared in a plugin manifest, so the prior approach (merging hooks into the project's .claude/settings.json) produced broken commands like /hooks/X.py (empty expansion). Hooks ship with the plugin from v6.13.0 onwards.

For existing installs, the old hook entries are duplicated in .claude/settings.json and must be removed to prevent double-firing:

bash
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/cache/navigator-marketplace/navigator}"
if [ ! -d "$PLUGIN_DIR" ]; then
  PLUGIN_DIR="$HOME/.claude/plugins/marketplaces/navigator-marketplace"
fi

python3 "$PLUGIN_DIR/skills/nav-upgrade/functions/migrate_hooks_out_of_settings.py" \
    .claude/settings.json

# Output:
#   ✓ Backup written → .claude/settings.json.pre-migrate.<timestamp>
#   ✓ Removed N Navigator hook entries (commands path matched)
#   ✓ Preserved K unrelated user hooks

Behavior:

  • Conservative match: command string must contain BOTH hooks/ AND one of the known Navigator hook names (nav_session_start, nav_pre_compact, nav_post_compact, nav_workflow_state, nav_read_guard, nav_task_graph_sync, nav_profile_sync, nav_commit_reminder, nav_brief, token_monitor, workflow_enforcer).
  • Writes a .pre-migrate.<timestamp> backup before any change.
  • Logs every removed entry to stderr.
  • Idempotent: re-running on an already-migrated file is a no-op.
  • Never touches entries with unrelated commands.

Fresh installs do not need this step — they never had hooks in .claude/settings.json to begin with.

v7.0.0 migration note: the nine per-hook scripts were replaced by a single dispatcher (hooks/nav_dispatch.py) routing to hooks/ops/* — the old files no longer exist in the plugin. Any consumer settings.json entry still referencing an old hook path (e.g. hooks/nav_session_start.py, hooks/workflow_enforcer.py, hooks/nav_brief.py) is stale and now fails with "No such file or directory" on every event. Run the same migrate_hooks_out_of_settings.py step above — its basename list covers all nine v6 hooks — then restart Claude Code. Per-hook state files are archived automatically by the SessionStart op to .agent/.nav-v6-state.bak/ (replaced by .agent/.nav-runtime-state.json, schema 2).

⚠️  RESTART REQUIRED to activate hooks from the plugin manifest.
   Claude Code caches plugin hook definitions at session start.

Step 6: Post-Upgrade Setup Check

Check if new features require setup:

bash
# Check for skills with setup requirements
if [ -f "$NAVIGATOR_PATH/skills/product-design/setup.sh" ]; then
  # Check if venv exists
  if [ ! -d "$NAVIGATOR_PATH/skills/product-design/venv" ]; then
    echo "⚠️  product-design skill requires setup"
    NEEDS_SETUP=true
  fi
fi

If setup needed, show instructions:

markdown
⚠️  New Feature Requires Setup

The product-design skill (v3.4.0+) requires Python dependencies:

**One-time setup** (30 seconds):
```bash
cd ~/.claude/plugins/marketplaces/navigator-marketplace/skills/product-design
./setup.sh

What this installs:

  • Python MCP SDK for direct Figma connection
  • 95% orchestration reduction
  • 92% token savings

After setup, use: "Review this Figma design: [URL]"


**Record setup needed in TodoWrite** for tracking.

---

### Step 7: Feature Discovery

**Show new features** available in updated version.

**For v3.3.0 update**:
````markdown
🎉 Navigator v3.3.0 Update Complete!

## New Features Available

### visual-regression Skill (NEW)
Set up Storybook + Chromatic in 5 minutes instead of 2-3 hours.

**Usage**:

"Set up visual regression for ProfileCard" "Add Chromatic to Button component" "Configure visual tests for Input, Card, Modal"


**What it does**:
✅ Generates Storybook stories with all variants
✅ Configures Chromatic/Percy/BackstopJS
✅ Creates CI workflows (GitHub Actions, GitLab CI)
✅ Adds accessibility tests

**Complete Design Pipeline** (v3.2 + v3.3):
1. "Review this design from Figma" (v3.2)
2. Implement components
3. "Set up visual regression" (v3.3 NEW)
4. Automated visual testing in CI

### Updated Skills Count
- **17 total skills** (was 16)
- 10 core Navigator skills
- 7 development skills

### Integration
visual-regression integrates with product-design skill for complete design→code→testing workflow.

## Try It Now

If you have Storybook in this project:

"Set up visual regression for [ComponentName]"


If you don't have Storybook:
```bash
npx storybook init

Then:

"Set up visual regression for [ComponentName]"

Documentation


---

## Predefined Functions

### functions/version_detector.py

**Purpose**: Detect current and latest Navigator versions

**Usage**:
```bash
python3 functions/version_detector.py
```

**Output**:
```json
{
  "current_version": "3.2.0",
  "latest_version": "3.3.0",
  "update_available": true,
  "release_url": "https://github.com/alekspetrov/navigator/releases/tag/v3.3.0",
  "changes": {
    "new_skills": ["visual-regression"],
    "updated_skills": ["product-design"],
    "new_features": ["Multi-tool VR support", "CI workflows"],
    "breaking_changes": []
  }
}
```

### functions/plugin_updater.py

**Purpose**: Execute plugin update with retry logic

**Usage**:
```bash
python3 functions/plugin_updater.py --target-version 3.3.0
```

**Actions**:
1. Execute `/plugin update navigator`
2. If fails, retry with uninstall/reinstall
3. Verify update succeeded
4. Return status

### functions/plugin_verifier.py

**Purpose**: Verify update completed successfully

**Usage**:
```bash
python3 functions/plugin_verifier.py --expected-version 3.3.0
```

**Checks**:
- Plugin version matches expected
- New skills exist in filesystem
- Skills registered in plugin.json
- Skills are invokable (test invocation)

---

## Error Handling

### Update Failed: Network Error

```
❌ Update failed: Could not connect to plugin marketplace

Fix:
1. Check internet connection
2. Try again in a few minutes
3. Manual update: /plugin uninstall navigator && /plugin install navigator
```

### Update Failed: Permission Denied

```
❌ Update failed: Permission denied

Fix:
1. Close Claude Code
2. Check ~/.config/claude/plugins/ permissions
3. Restart Claude Code
4. Try update again
```

### Verification Failed: Skills Not Found

```
⚠️ Update completed but new skills not found

Fix:
1. Restart Claude Code (required for skill reload)
2. Verify: /plugin list
3. Test: "Set up visual regression for Button"
```

Automatically prompt user to restart.

### CLAUDE.md Update Conflicts

```
⚠️ CLAUDE.md update has conflicts with your customizations

Options:
[1] Keep my customizations (merge new features)
[2] Use new template (lose customizations)
[3] Show me the diff first

Reply with choice
```

Let user decide how to handle conflicts.

---

## Upgrade Paths

### From v3.0.x to v3.3.0

**Changes**:
- +2 skills (nav-markers in v3.1, visual-regression in v3.3)
- OpenTelemetry integration (v3.1)
- Product design skill (v3.2)
- Visual regression skill (v3.3)

**Breaking changes**: None (fully backward compatible)

### From v3.1.x to v3.3.0

**Changes**:
- Product design skill (v3.2)
- Visual regression skill (v3.3)
- Updated skills count (17 total)

**Breaking changes**: None

### From v3.2.x to v3.3.0

**Changes**:
- Visual regression skill
- Integration with product-design workflow
- Updated skills count (17 total)

**Breaking changes**: None

---

## Post-Update Checklist

After upgrade, verify:

- ✅ `/plugin list` shows new version
- ✅ CLAUDE.md updated with new patterns
- ✅ New skills auto-invoke on natural language
- ✅ Existing skills still work
- ✅ No conflicts in project configuration

**If all checked**: Update successful!

---

## Rollback

If update causes issues:

```
"Rollback Navigator to v3.2.0"
```

This will:
1. Uninstall current version
2. Install specific version from marketplace
3. Update CLAUDE.md to match
4. Verify rollback succeeded

---

## Integration Points

### With nav-sync-claude

After plugin update, automatically invokes `nav-sync-claude` to sync project configuration.

### With nav-start

After update, `nav-start` shows new features available in session statistics.

### With nav-init

If upgrading before project initialization, suggests running `nav-init` with latest features.

---

## Examples

### Example 1: Simple Update

```
User: "Update Navigator"

→ Detects: v3.2.0 → v3.3.0 available
→ Updates plugin
→ Updates CLAUDE.md
→ Shows: "visual-regression skill now available"
→ Suggests: "Set up visual regression for [Component]"
```

### Example 2: Already on Latest

```
User: "Update Navigator"

→ Detects: Already on v3.3.0
→ Shows new features available
→ Suggests trying visual-regression if not used yet
```

### Example 3: Update with Restart Required

```
User: "Update Navigator"

→ Updates plugin
→ Verification: Skills not found (needs restart)
→ Prompts: "Please restart Claude Code to complete update"
→ After restart: Verification succeeds
```

---

## Best Practices

1. **Update regularly**: Check for updates monthly
2. **Read release notes**: Understand new features before using
3. **Test new skills**: Try new features in test project first
4. **Report issues**: File GitHub issues for update problems
5. **Backup CLAUDE.md**: Keep backup before update (auto-created)

---

## Version History

- **v1.0.0**: Initial nav-upgrade skill (Navigator v3.3.1)

---

## Future Enhancements

- Auto-update check on `nav-start` (opt-in)
- Changelog display in CLI
- Update notifications for major versions
- Automated migration scripts for breaking changes

---

**Last Updated**: 2025-10-21
**Skill Type**: Core Navigator
**Auto-Invocation**: Yes

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Nav Upgrade AI skill do?

Automates Navigator plugin updates. Detects current version, updates plugin, verifies installation, updates project CLAUDE.md, and validates new features. Auto-invoke when user mentions upgrading Navigator or getting new features.

Why use Nav Upgrade on TypingMind?

Because you install it once and use it with any model. Nav Upgrade is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Nav Upgrade in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/qf-studio/navigator/tree/main/skills/nav-upgrade. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Nav Upgrade?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Nav Upgrade?

As many as you like. As long as a model supports skills, you can use Nav Upgrade with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Nav Upgrade AI skill free?

Yes. It is published on GitHub by qf-studio under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇