What does the Nav Stats AI skill do?

Display session efficiency report showing token savings, cache performance, and optimization recommendations. Use when user asks "show my stats", "how efficient am I?", "show session metrics", or wants to see Navigator's impact.

Why use Nav Stats on TypingMind?

Because you install it once and use it with any model. Nav Stats 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 Stats in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/qf-studio/navigator/tree/main/skills/nav-stats. 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 Stats?

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 Stats?

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

Is the Nav Stats 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.

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.

Nav Stats by qf-studio — AI Skill for Any LLM

Navigator Session Statistics Skill

Show real-time efficiency reporting with baseline comparisons, making Navigator's value quantifiable and shareable.

When to Invoke

Invoke this skill when the user:

Says "show my stats", "show session stats", "show metrics"
Asks "how efficient am I?", "how much did I save?"
Says "show my Navigator report", "efficiency report"
Wants to see token savings or session performance
Says "show impact", "prove Navigator works"

DO NOT invoke if:

User just started session (< 5 messages)
Navigator not initialized in project
User asking about specific metrics only (answer directly)

Execution Steps

Step 1: Check Navigator Initialized

Verify Navigator is set up:

bash
if [ ! -f ".agent/DEVELOPMENT-README.md" ]; then
  echo "❌ Navigator not initialized in this project"
  echo "Run 'Initialize Navigator' first"
  exit 1
fi

Step 2: Run Enhanced Session Stats

Execute the enhanced session statistics script:

bash
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/cache/navigator-marketplace/navigator}"
[ -d "$PLUGIN_DIR" ] || PLUGIN_DIR="$HOME/.claude/plugins/marketplaces/navigator-marketplace"

# Check if enhanced script exists
if [ ! -f "$PLUGIN_DIR/scripts/session-stats.sh" ]; then
  echo "❌ Session stats script not found"
  echo "Reinstall or update Navigator to restore scripts/session-stats.sh"
  exit 1
fi

# Run stats script
bash "$PLUGIN_DIR/scripts/session-stats.sh"

This script outputs shell-parseable variables:

BASELINE_TOKENS - Total size of all .agent/ docs
LOADED_TOKENS - Actually loaded in session (estimated)
TOKENS_SAVED - Difference
SAVINGS_PERCENT - Percentage saved
EFFICIENCY_SCORE - 0-100 score
CACHE_EFFICIENCY - From OpenTelemetry
CONTEXT_USAGE_PERCENT - Estimated context fill
TIME_SAVED_MINUTES - Estimated time saved

Step 3: Calculate Efficiency Score

Use predefined function to calculate score:

bash
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/cache/navigator-marketplace/navigator}"
[ -d "$PLUGIN_DIR" ] || PLUGIN_DIR="$HOME/.claude/plugins/marketplaces/navigator-marketplace"

# Extract metrics from session-stats.sh
source <(bash "$PLUGIN_DIR/scripts/session-stats.sh")

# Calculate efficiency score using predefined function
EFFICIENCY_SCORE=$(python3 "$PLUGIN_DIR/skills/nav-stats/functions/efficiency_scorer.py" \
  --tokens-saved-percent ${SAVINGS_PERCENT} \
  --cache-efficiency ${CACHE_EFFICIENCY} \
  --context-usage ${CONTEXT_USAGE_PERCENT})

Step 4: Format and Display Report

Use predefined function to format visual report:

bash
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/cache/navigator-marketplace/navigator}"
[ -d "$PLUGIN_DIR" ] || PLUGIN_DIR="$HOME/.claude/plugins/marketplaces/navigator-marketplace"

# Generate formatted report
python3 "$PLUGIN_DIR/skills/nav-stats/functions/report_formatter.py" \
  --baseline ${BASELINE_TOKENS} \
  --loaded ${LOADED_TOKENS} \
  --saved ${TOKENS_SAVED} \
  --savings-percent ${SAVINGS_PERCENT} \
  --cache-efficiency ${CACHE_EFFICIENCY} \
  --context-usage ${CONTEXT_USAGE_PERCENT} \
  --efficiency-score ${EFFICIENCY_SCORE} \
  --time-saved ${TIME_SAVED_MINUTES}

Output Format:


╔══════════════════════════════════════════════════════╗
║          NAVIGATOR EFFICIENCY REPORT                 ║
╚══════════════════════════════════════════════════════╝

📊 TOKEN USAGE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Documentation loaded:        12,000 tokens
Baseline (all docs):        150,000 tokens
Tokens saved:               138,000 tokens (92% ↓)

💾 CACHE PERFORMANCE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Cache efficiency:              100.0% (perfect)

📈 SESSION METRICS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Context usage:                      35% (excellent)
Efficiency score:                94/100 (excellent)

⏱️  TIME SAVED
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Estimated time saved:          ~42 minutes

💡 WHAT THIS MEANS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Navigator loaded 92% fewer tokens than loading all docs.
Your context window is 65% available for actual work.

🎯 RECOMMENDATIONS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
✅ Excellent efficiency - keep using lazy-loading strategy
✅ Context usage healthy - plenty of room for work

Share your efficiency: Take a screenshot! #ContextEfficiency

Step 4.5: Tier-1 Telemetry Row (v7.0.0+)

If .agent/.nav-runtime-state.json (schema 2) carries a tier1 section, append one row to the report:


Tier-1 responder: {hits} zero-token answers | {false_positives} suspected false positives

A false positive = a Tier-1 hit followed by a near-identical re-prompt (the user wanted the model after all). Rising false positives mean the exact-match table is intercepting prompts it should not — suggest disabling the offending rule via tier1.rules.<id>: false. Omit the row when the section is absent or tier1 is disabled.

Step 5: Add Context-Specific Recommendations

Based on efficiency score, provide actionable advice:

If efficiency_score < 70:


⚠️  RECOMMENDATIONS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚠️  Token savings below target (70%+)
→ Check: Are you loading more docs than needed?
→ Tip: Use navigator to find docs, don't load all upfront

Read more: .agent/philosophy/CONTEXT-EFFICIENCY.md

If context_usage > 80%:


⚠️  RECOMMENDATIONS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚠️  Context usage high (80%+)
→ Consider: Create context marker and compact
→ Tip: Compact after completing sub-tasks

Read more: .agent/philosophy/ANTI-PATTERNS.md

If cache_efficiency < 80%:


⚠️  RECOMMENDATIONS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚠️  Cache efficiency low (<80%)
→ Check: CLAUDE.md properly configured?
→ Tip: Ensure prompt caching enabled

Read more: .agent/philosophy/PATTERNS.md (Caching pattern)

Predefined Functions

`efficiency_scorer.py`

Calculate Navigator efficiency score (0-100) based on:

Token savings (40 points)
Cache efficiency (30 points)
Context usage (30 points)

Usage:

bash
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/cache/navigator-marketplace/navigator}"
[ -d "$PLUGIN_DIR" ] || PLUGIN_DIR="$HOME/.claude/plugins/marketplaces/navigator-marketplace"

python3 "$PLUGIN_DIR/skills/nav-stats/functions/efficiency_scorer.py" \
  --tokens-saved-percent 92 \
  --cache-efficiency 100 \
  --context-usage 35

Output: 94 (integer score)

`report_formatter.py`

Format efficiency metrics into visual, shareable report.

Usage:

bash
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/cache/navigator-marketplace/navigator}"
[ -d "$PLUGIN_DIR" ] || PLUGIN_DIR="$HOME/.claude/plugins/marketplaces/navigator-marketplace"

python3 "$PLUGIN_DIR/skills/nav-stats/functions/report_formatter.py" \
  --baseline 150000 \
  --loaded 12000 \
  --saved 138000 \
  --savings-percent 92 \
  --cache-efficiency 100 \
  --context-usage 35 \
  --efficiency-score 94 \
  --time-saved 42

Output: Formatted ASCII report (see Step 4)

Philosophy Integration

Context Engineering Principle: Measurement validates optimization

From .agent/philosophy/PATTERNS.md:

"Measure to validate. Navigator tracks real metrics, not estimates."

This skill proves:

Token savings are real (baseline comparison)
Cache efficiency works (OpenTelemetry data)
Context usage is healthy (window not overloaded)
Time saved is quantifiable (6s per 1k tokens)

User Experience

User says: "Show my stats"

Skill displays:

Visual efficiency report
Clear metrics (tokens, cache, context)
Interpretation ("What this means")
Actionable recommendations

User can:

Screenshot and share (#ContextEfficiency)
Understand Navigator's impact
Optimize workflow based on recommendations
Validate context engineering principles

Example Output Scenarios

Scenario 1: Excellent Efficiency (Score 94)

User following lazy-loading pattern, cache working perfectly:

92% token savings ✅
100% cache efficiency ✅
35% context usage ✅
Score: 94/100

Recommendation: Keep it up! Share your efficiency.

Scenario 2: Fair Efficiency (Score 72)

User loading too many docs upfront:

65% token savings ⚠️
95% cache efficiency ✅
55% context usage ✅
Score: 72/100

Recommendation: Review lazy-loading strategy. Load docs on-demand.

Scenario 3: Poor Efficiency (Score 48)

User not using Navigator patterns:

45% token savings ❌
70% cache efficiency ⚠️
85% context usage ❌
Score: 48/100

Recommendation: Read philosophy docs. Consider /nav:compact. Review CLAUDE.md.

Success Metrics

After using this skill, users should:

Understand their efficiency score
See quantified token savings
Know what to improve (if anything)
Feel motivated to share results

Long-term impact:

Users screenshot reports and share
"Navigator saved me 138k tokens" becomes common
Efficiency becomes visible, not abstract
Continuous improvement through measurement

This skill makes Navigator's value tangible and shareable.

Publisher	qf-studio
Repository	`navigator`
Skill name	`nav-stats`
Stars	232
Forks	12
Bundled files	2
License	MIT
Links	Skill source Repository Homepage

Nav Stats

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Navigator Session Statistics Skill

When to Invoke

Execution Steps

Step 1: Check Navigator Initialized

Step 2: Run Enhanced Session Stats

Step 3: Calculate Efficiency Score

Step 4: Format and Display Report

Step 4.5: Tier-1 Telemetry Row (v7.0.0+)

Step 5: Add Context-Specific Recommendations

Predefined Functions

`efficiency_scorer.py`

`report_formatter.py`

Philosophy Integration

User Experience

Example Output Scenarios

Scenario 1: Excellent Efficiency (Score 94)

Scenario 2: Fair Efficiency (Score 72)

Scenario 3: Poor Efficiency (Score 48)

Success Metrics

Bundled files

Frequently asked questions

Set up your own AI workspace now

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Navigator Session Statistics Skill

When to Invoke

Execution Steps

Step 1: Check Navigator Initialized

Step 2: Run Enhanced Session Stats

Step 3: Calculate Efficiency Score

Step 4: Format and Display Report

Step 4.5: Tier-1 Telemetry Row (v7.0.0+)

Step 5: Add Context-Specific Recommendations

Predefined Functions

efficiency_scorer.py

report_formatter.py

Philosophy Integration

User Experience

Example Output Scenarios

Scenario 1: Excellent Efficiency (Score 94)

Scenario 2: Fair Efficiency (Score 72)

Scenario 3: Poor Efficiency (Score 48)

Success Metrics

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now

`efficiency_scorer.py`

`report_formatter.py`