skills.rest
qf-studio
qf-studioOFFICIAL
Explore all skills by this author →
WebsiteGitHub

nav-stats

Generate session efficiency reports showing token savings, cache performance, and optimization recommendations.

232|12|Updated Oct 10, 2025
Search Tags:#claude code#context engineering#metrics#token savings#cache performance#session stats#efficiency report
One-click install
npx skills add https://github.com/qf-studio/navigator --skill nav-stats-qf-studio
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: nav-stats
Source: https://github.com/qf-studio/navigator/tree/main/skills/nav-stats
Command: npx skills add https://github.com/qf-studio/navigator --skill nav-stats-qf-studio

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Developers using context engineering lack visibility into whether lazy-loading documentation actually saves tokens and keeps sessions healthy, making optimization efforts unmeasurable. ## Core Features & Use Cases - Efficiency Reporting: Runs session-stats.sh to compute baseline vs loaded tokens, savings percentage, cache efficiency, and context usage. - Scoring & Formatting: Calculates a 0-100 efficiency score via efficiency_scorer.py and renders a shareable ASCII report via report_formatter.py. - Actionable Recommendations: Surfaces targeted advice when token savings drop below 70%, cache efficiency falls under 80%, or context usage exceeds 80%. - Use Case: After a long coding session, ask for your stats to see that you saved 138k tokens (92%) with a 94/100 efficiency score, then screenshot the report to share. ## Quick Start Ask the assistant to show your Navigator session stats and efficiency report.

Frequently Asked Questions about nav-stats

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I see token savings from my Claude Code session?▼

Ask to show your session stats and the skill runs session-stats.sh to compare baseline documentation size against actually loaded tokens. It displays savings percentage, an efficiency score, and estimated time saved in a formatted report.

How is the Navigator efficiency score calculated?▼

The score combines three weighted metrics: token savings worth up to 40 points, cache efficiency up to 30 points, and context usage up to 30 points. Scores of 90 or above are rated excellent, while below 70 is poor.

Why does the stats skill say Navigator is not initialized?▼

The skill requires a .agent/DEVELOPMENT-README.md file in the project to confirm Navigator setup. If missing, it exits with an error and instructs you to initialize Navigator first.

What should I do when my efficiency score is low?▼

The report provides targeted recommendations based on which metric is weak. Low token savings suggests loading fewer docs upfront, low cache efficiency suggests checking CLAUDE.md caching configuration, and high context usage suggests compacting after sub-tasks.

Can I see Tier-1 responder statistics in the report?▼

Yes, if .agent/.nav-runtime-state.json contains a tier1 section, the report appends a row showing zero-token answers and suspected false positives. Rising false positives suggest disabling the offending rule.

Related Skills

Explore More Software Engineering Skills→
ruifrvazruifrvaz
3

smaqit.session-start

Loads project files and operational context to prepare AI for answering user queries.

Community
Intermediate
pagerguildpagerguild

context-loader

Load tiered project context for cross-session and cross-agent workflows.

Official
Advanced
TykoDevTykoDev
9

quality-review

Analyze code health by measuring maintainability, standards conformance, and architectural drift.

Community
Advanced
aaronachermannaaronachermann

querying-mlflow-metrics

Fetch aggregated MLflow trace metrics with time-range bucketing and JSON or table output.

Community
Intermediate