Is Naval safe to install?

SaferSkills independently audited Naval (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.

CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

100

35.0 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 0 flagged

Securityscore 100 · 0 findings

✓—All security checks passedNo findings in this category for the latest scan.pass

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

README.md · 1 file

README.md26.1 KB · Markdown

Unofficial Naval Skills

naval is an unofficial plugin and skill pack for applying ideas from The Almanack of Naval Ravikant.

It gives agents 79 searchable n-* skills for wealth, judgment, happiness, health, values, reading, decisions, scorecards, daily reviews, weekly reviews, optional memory, source fidelity, and quote safety.

This repository does not include the full book text. It contains source-safe skill instructions, workflows, and reference maps.

It is not affiliated with Naval Ravikant, Eric Jorgenson, Navalmanack, or any publisher.

What The Skills Do

Each skill gives the agent a specific job. The pack is split this way so the agent can load the right instructions for the user's situation instead of one giant prompt.

Skill Type	What It Does	Examples
Router	Chooses the best Naval skill when the request is vague.	`n-router`
Wealth skills	Apply ideas about specific knowledge, accountability, ownership, leverage, long-term games, and risk.	`n-wealth-map`, `n-specific-knowledge`, `n-leverage-stack`
Judgment skills	Improve decisions with clear thinking, inversion, mental models, truth tests, and long-term consequences.	`n-decision-rules`, `n-clear-thinking`, `n-mental-models`
Happiness skills	Work through desire, envy, peace, presence, acceptance, and happiness habits.	`n-desire-audit`, `n-envy-antidote`, `n-presence-practice`
Health and agency skills	Build practical systems for health, exercise, diet, meditation, habits, and attention.	`n-health-first`, `n-meditation-system`, `n-habit-change`
Reading and philosophy skills	Build reading paths, values filters, meaning checks, and source trails.	`n-reading-curriculum`, `n-values-filter`, `n-next-sources`
Review and safety skills	Run daily/weekly reviews, score opportunities, check relationships, and keep quote use safe.	`n-weekly-compound-review`, `n-opportunity-scorecard`, `n-quote-safety`
Memory skills	Configure optional local output, save approved learnings, and refresh stale saved guidance.	`n-setup`, `n-save-learning`, `n-memory-refresh`

The Loop

n-router
  -> picks the best skill for vague "use Naval on this" requests

n-wealth-map / n-decision-rules / n-desire-audit / n-health-first / ...
  -> applies one focused lens to the situation

n-principle-to-action / n-opportunity-scorecard / n-relationship-scorecard
  -> turns insight into an experiment, scorecard, boundary, or next action

n-daily-review / n-weekly-compound-review
  -> checks whether the actions are compounding

n-setup / n-save-learning / n-memory-refresh
  -> optionally saves approved learnings and keeps local memory clean

n-source-fidelity / n-quote-safety / n-coverage-auditor
  -> keeps the pack faithful, safe, and complete

Use whichever step fits the moment. If the user already knows the problem is about desire, start with n-desire-audit. If the situation is vague, start with n-router.

Core Workflows

Workflow	Start With	Then Use	Produces
Route a vague situation	`n-router`	The routed primary skill plus 1-3 secondary lenses.	Skill path, first question, output format.
Build a wealth path	`n-wealth-map`	`n-specific-knowledge`, `n-leverage-stack`, `n-equity-ownership`, `n-patience-compounding`.	Wealth strategy, leverage moves, ownership experiments.
Make a hard decision	`n-decision-rules`	`n-inversion-filter`, `n-risk-of-ruin`, `n-wisdom-long-term`.	Decision frame, avoid list, reversible next step.
Evaluate an opportunity	`n-opportunity-scorecard`	`n-long-term-games`, `n-accountability-risk`, `n-work-as-play`.	Scorecard, red flags, test plan.
Debug unhappiness	`n-desire-audit`	`n-envy-antidote`, `n-accept-change-leave`, `n-presence-practice`.	Desire map, cost, replacement practice.
Rebuild health and agency	`n-health-first`	`n-exercise-priority`, `n-diet-simplifier`, `n-meditation-system`, `n-habit-change`.	Health baseline, habit system, daily practice.
Learn and read better	`n-reading-curriculum`	`n-reading-system`, `n-foundational-learning`, `n-next-sources`.	Reading path, drop rules, weekly output.
Compound the week	`n-weekly-compound-review`	Any skills surfaced by the review.	Wins, leaks, deletion, commitment, review date.
Set up optional memory	`n-setup`	`n-save-learning`, `n-memory-refresh`.	Local config, output folders, schemas, approved saved learnings.

Quick Example

Use n-router when you do not know the exact skill name:

Use Naval to analyze whether I should leave my job for a startup idea.

The router should return a path like:

Step	Skill	Why
1	`n-big-life-decisions`	Frames the job, city, company, and life-direction tradeoff.
2	`n-opportunity-scorecard`	Scores the startup idea on leverage, upside, compounding, and risk.
3	`n-risk-of-ruin`	Looks for health, financial, legal, relationship, and reputation blowups.
4	`n-principle-to-action`	Turns the decision into a reversible experiment or clear commitment.

For more walkthroughs, see docs/EXAMPLES.md.

Getting Started

Clone the repo:

mkdir -p ~/plugins
git clone https://github.com/thepraggyverse/naval.git ~/plugins/naval
cd ~/plugins/naval

Validate it:

python3 scripts/generate_skill_docs.py
python3 scripts/validate_public.py
python3 scripts/check_coverage.py
python3 scripts/validate_direct_install.py
python3 scripts/smoke_install.py

python3 scripts/install_local.py --marketplace

That makes the plugin available to Codex. To install and enable it from the Codex CLI:

codex plugin add naval@personal --json
codex plugin list | grep 'naval@personal'

Optionally expose every skill to local agent homes:

python3 scripts/install_local.py --symlink-skills

For Codex-only use, the plugin install is enough. Use direct skill symlinks for tools that scan SKILL.md folders, and prefer --skill-home when you only need one harness.

For both:

python3 scripts/install_local.py --marketplace --symlink-skills

Then search for n- in your agent or Codex skill picker.

Quick live smoke after installing:

python3 scripts/smoke_install.py --codex --live

Optional Memory Layer

naval is stateless by default. It does not save private reflections, reviews, or decisions unless the user explicitly opts in.

Use n-setup when you want durable outputs. It can configure no persistence, project-local docs/naval/, a personal notes or Obsidian folder, or a custom path. Saved artifacts use small schemas and templates under references/memory/, and every write should be confirmed before it happens.

Useful follow-up skills:

Skill	Use When	Saves To
`n-save-learning`	A session produced 1-3 reusable insights worth keeping.	`learnings/`
`n-memory-refresh`	Saved guidance may be stale, duplicated, contradictory, or low-value.	Updates existing memory files after approval.
`n-daily-review`	You want a dated review log after a daily review.	`reviews/`
`n-weekly-compound-review`	You want a dated weekly compounding review.	`reviews/`

See docs/NAVAL_MEMORY.md for the privacy contract, folder layout, direct-copy install requirements, and memory templates.

Harness Support

Harness	Best Path	Details
Codex App / CLI	Native Codex plugin or local marketplace	docs/INSTALL.md
Claude Code	Claude plugin marketplace metadata	docs/HARNESS_SUPPORT.md
Cursor	Cursor plugin metadata or direct skills	docs/HARNESS_SUPPORT.md
GitHub Copilot	Source plugin or direct skills	docs/HARNESS_SUPPORT.md
Gemini CLI	Native Gemini extension	docs/HARNESS_SUPPORT.md
OpenCode	OpenCode plugin entrypoint	.opencode/INSTALL.md
Pi	Pi extension metadata	docs/HARNESS_SUPPORT.md
Kiro, OpenClaw, other SKILL.md loaders	Direct `SKILL.md` symlinks or copies	docs/SYMLINKS.md

Components

Type	Count	What It Does	Where
Harness metadata	multiple	Makes the pack visible to plugin-capable harnesses.	`.codex-plugin/`, `.claude-plugin/`, `.cursor-plugin/`, `.agents/plugins/`, `gemini-extension.json`, `.opencode/`, `.pi/`, `package.json`
Skills	79	Small callable `n-*` behaviors for applying book concepts.	skills/
Router	1	Routes vague requests to the right primary and secondary skills.	skills/n-router/SKILL.md
Catalog	1	Lists every skill with area, use case, and example prompt.	references/skill-catalog.md
Skill docs	79	Browsable GitHub pages for each skill.	docs/skills/
Coverage matrix	1	Maps book sections to skills so gaps are visible.	references/coverage-matrix.yaml
Workflows	10	Reusable scorecards and review protocols.	references/workflows/
Memory schemas/templates	14	Optional saved-output contracts for reviews, decisions, scorecards, experiments, practices, learnings, and quote notes.	references/memory/
Chapter summaries	7	Paraphrased coverage anchors by book area.	references/chapter-summaries/
Discovery metadata	1	Groups the skills for skills.sh-style browsers.	skills.sh.json
Installer	1	Updates the local marketplace and/or symlinks skills.	scripts/install_local.py
Direct-copy exporter	1	Builds a portable `skills/` plus `references/` bundle for SKILL.md-only hosts.	scripts/export_direct_install.py
Validators and smoke	4	Check public structure, skill count, references, coverage, copied-skill portability, generated docs, and install surfaces.	scripts/validate_public.py, scripts/check_coverage.py, scripts/validate_direct_install.py, scripts/smoke_install.py

Skill Areas

Area	Skills	Use For
Wealth	20	Specific knowledge, accountability, ownership, leverage, work as play, patience, reputation, risk, and freedom.
Judgment	10	Clear thinking, reality checks, decision rules, mental models, inversion, falsifiability, and learning foundations.
Happiness	9	Desire, peace, presence, envy, success games, acceptance, habits, and death awareness.
Saving Yourself	15	Health, exercise, diet, meditation, habit change, systems, anger, expectations, and modern addiction defense.
Philosophy	5	Meaning, values, rational inner work, long-term wisdom, and present action.
Reading	4	Reading systems, curricula, formulas, rules, and source trails.
Meta	14	Routing, setup, memory, coverage, quote safety, daily/weekly reviews, scorecards, coaching, flashcards, and book clubs.
Front Matter	2	Source fidelity and biographical context.

Full Skill Inventory

Every callable skill uses the n- prefix so search can find it quickly.

Skill	Area	Use When
`n-router`	meta	You want Naval applied but do not know which skill to start with.
`n-setup`	meta	You want to configure optional output folders, privacy defaults, and direct-install reference guidance.
`n-source-fidelity`	front-matter	You need faithful interpretation, attribution, or a citation-safety check.
`n-biographical-context`	front-matter	You want life context for why a principle appears in the book.
`n-wealth-map`	wealth	You want an end-to-end wealth strategy.
`n-wealth-vs-money-status`	wealth	You need to separate wealth creation from money, approval, prestige, or hierarchy.
`n-specific-knowledge`	wealth	You want to identify hard-to-train strengths and authentic edges.
`n-authenticity-positioning`	wealth	You want positioning that is natural, rare, and hard to copy.
`n-long-term-games`	wealth	You need to evaluate compounding relationships, markets, or work.
`n-integrity-partner-filter`	wealth	You need to assess a collaborator or partner.
`n-accountability-risk`	wealth	You are deciding whether to put your name or reputation behind something.
`n-equity-ownership`	wealth	You want to move from wages toward ownership, IP, assets, or upside.
`n-leverage-stack`	wealth	You want to add code, media, capital, labor, brand, distribution, or automation leverage.
`n-build-or-sell`	wealth	You are choosing whether to learn building, selling, or both.
`n-judgment-compensation`	wealth	You want to become paid for judgment under leverage, not just time.
`n-time-value-focus`	wealth	You need to clear the calendar and apply personal hourly-rate thinking.
`n-status-game-detector`	wealth	You suspect zero-sum approval, politics, prestige, or comparison is driving behavior.
`n-big-life-decisions`	wealth	You are making a job, city, company, partner, or life-direction decision.
`n-work-as-play`	wealth	You want work that feels like play to you and work to others.
`n-retirement-freedom-design`	wealth	You want a path to freedom through income, low burn, or loved work.
`n-luck-engineering`	wealth	You want to increase luck through motion, skill, reputation, and character.
`n-reputation-networking`	wealth	You want relationships built through craft and generosity instead of shallow networking.
`n-patience-compounding`	wealth	You need to balance urgent action with long-term compounding.
`n-risk-of-ruin`	wealth	You need to find legal, health, financial, relationship, or reputation ruin risks.
`n-judgment-builder`	judgment	You want better decision quality in leveraged work or life choices.
`n-clear-thinking`	judgment	You want confused thinking simplified to basics and plain language.
`n-reality-ego-audit`	judgment	You need to see where desire, ego, or fear blocks reality.
`n-identity-shedding`	judgment	You want to drop inherited labels, tribes, packages, or stale self-images.
`n-decision-rules`	judgment	You want compact decision heuristics applied to a choice.
`n-mental-models`	judgment	You want lenses like compounding, incentives, game theory, inversion, or probability.
`n-inversion-filter`	judgment	You need to find what fails, what to avoid, and which mistakes matter.
`n-falsifiability-truth`	judgment	You need to test whether a claim has predictive power.
`n-foundational-learning`	judgment	You want foundations in math, science, economics, persuasion, or learning.
`n-reading-system`	judgment	You want curiosity-led reading habits, rereading, skimming, and dropping weak books.
`n-happiness-baseline`	happiness	You want to train happiness as a baseline.
`n-desire-audit`	happiness	You need to identify a desire creating suffering.
`n-presence-practice`	happiness	You need to reduce rumination and return attention to present reality.
`n-peace-over-joy`	happiness	You want stable peace instead of stimulation or mood chasing.
`n-success-game-exit`	happiness	You may have outgrown a game with large rewards.
`n-envy-antidote`	happiness	You want to dissolve jealousy, comparison, or external scorecards.
`n-happiness-habits`	happiness	You want daily habits that support mood stability and happiness.
`n-accept-change-leave`	happiness	You need to choose acceptance, change, or exit.
`n-death-awareness`	happiness	You want mortality to clarify priorities and reduce ego battles.
`n-self-responsibility`	saving-yourself	You are outsourcing responsibility to gurus, mentors, doctors, teachers, or tools.
`n-be-yourself`	saving-yourself	You want original contribution instead of copying others.
`n-health-first`	saving-yourself	You need to reorder life around physical, mental, and spiritual health.
`n-diet-simplifier`	saving-yourself	You want to subtract processed food and appetite traps.
`n-exercise-priority`	saving-yourself	You need daily exercise to become non-negotiable.
`n-meditation-system`	saving-yourself	You want a practical awareness, sitting, walking, or journaling protocol.
`n-mental-debugging`	saving-yourself	You want to observe thoughts, reactions, fear, desire, and ego as processes.
`n-habit-change`	saving-yourself	You want triggers, substitutes, tracking, and commitment for habit change.
`n-systems-not-goals`	saving-yourself	You want to turn a goal into an environment or system.
`n-grow-yourself`	saving-yourself	You want self-change through self-image, learning, and readiness.
`n-freedom-from-expectations`	saving-yourself	You need boundaries around obligations and others' imagined claims.
`n-anger-release`	saving-yourself	You want a cleaner response to anger.
`n-employment-freedom`	saving-yourself	You want to reduce dependence on employment.
`n-uncontrolled-thinking`	saving-yourself	You want the mind to become a tool instead of the master.
`n-modern-addiction-defense`	saving-yourself	You need to audit screens, news, processed food, games, porn, stimulants, or dopamine traps.
`n-meaning-maker`	philosophy	You want to explore personal meaning without borrowing a universal answer.
`n-values-filter`	philosophy	You need to check honesty, long-term thinking, peer relationships, anger, or freedom.
`n-rational-buddhism`	philosophy	You want testable inner work without unverifiable claims as operating truth.
`n-wisdom-long-term`	philosophy	You want to evaluate long-term consequences.
`n-present-moment`	philosophy	You need to return to the present and act on perishable inspiration.
`n-reading-curriculum`	reading	You want book paths across science, mental models, philosophy, economics, and fiction.
`n-life-formulas`	reading	You want formula-style framing for health, wealth, happiness, income, or learning.
`n-rules`	reading	You want compact rules as reminders, not substitutes for reasoning.
`n-next-sources`	reading	You want deeper Naval resources and source trails.
`n-coverage-auditor`	meta	You want to check whether the plugin missed a major book section.
`n-principle-to-action`	meta	You want to turn a principle into an experiment, behavior, or operating rule.
`n-save-learning`	meta	You want to save 1-3 approved reusable insights from a completed Naval session.
`n-memory-refresh`	meta	You want to review saved Naval memory for stale, duplicate, contradictory, or low-value entries.
`n-daily-review`	meta	You want a daily health, work, desire, focus, and freedom review.
`n-weekly-compound-review`	meta	You want a weekly review across wealth, judgment, health, happiness, relationships, and values.
`n-opportunity-scorecard`	meta	You want to score a job, startup, product, project, content idea, or investment.
`n-relationship-scorecard`	meta	You want to score collaborators, friends, partners, or teams for long-term fit.
`n-quote-safety`	meta	You want quotations kept short, attributed, and compliant.
`n-socratic-coach`	meta	You want questions before advice.
`n-flashcards`	meta	You want recall prompts, reflection cards, or spaced repetition items.
`n-book-club`	meta	You want a chapter discussion, exercise, or reading reflection.

Common Prompts

Prompt	Start With	Good Output Shape
"Use Naval to analyze this career decision."	`n-router`	Route, first question, decision scorecard, next action.
"Build me a Naval-style wealth plan."	`n-wealth-map`	Wealth map, leverage stack, ownership moves, patient experiments.
"Help me identify my specific knowledge."	`n-specific-knowledge`	Strength inventory, obsession map, marketable edge, practice loop.
"Apply Naval's decision rules to this choice."	`n-decision-rules`	One-way/two-way door check, inversion, long-term consequences.
"What desire is making me unhappy?"	`n-desire-audit`	Desire, cost, root, accept/change/leave path, replacement practice.
"Run my weekly compound review."	`n-weekly-compound-review`	Review table, compounding wins, leaks, next week constraints.
"Set up Naval memory for this repo."	`n-setup`	Storage mode, privacy default, config draft, folder plan.
"Save what we learned from this Naval session."	`n-save-learning`	1-3 approved learnings with YAML frontmatter.
"Refresh my saved Naval learnings."	`n-memory-refresh`	Keep/update/merge/delete recommendations.
"Build me a Naval reading curriculum."	`n-reading-curriculum`	Reading path, sequence, reread list, drop criteria.

Install Paths

Path	Use When	Command
Codex plugin	You want the plugin visible as `naval` in Codex.	`python3 scripts/install_local.py --marketplace`
Direct skill symlinks	Your agent scans `SKILL.md` folders directly.	`python3 scripts/install_local.py --symlink-skills`
Both	You use multiple local agents and want one source of truth.	`python3 scripts/install_local.py --marketplace --symlink-skills`
Dry run	You want to inspect changes before writing.	`python3 scripts/install_local.py --marketplace --symlink-skills --dry-run`

See docs/INSTALL.md and docs/SYMLINKS.md for full install details.

How It Fits Together

Question	Answer
Is this a plugin or skills?	Both. The plugin is the installable package; the `n-*` skills are the callable units.
Why not one giant skill?	Smaller skills route better, load less context, and make each workflow easier to inspect.
Why `n-`?	It makes every skill searchable by a short namespace.
Where is the full list?	references/skill-catalog.md is the canonical catalog.
How do I check coverage?	Run `python3 scripts/check_coverage.py` or call `n-coverage-auditor`.
Where do saved outputs go?	Run `n-setup`; by default memory is off, and project-local memory uses `docs/naval/`.
Do copied skills need references?	Yes. Use `npm run export:direct` or copy `skills/n-*` and the sibling `references/` folder together; validate with `python3 scripts/validate_direct_install.py --agent-root <agent-root>`.
Can I use exact quotes?	Use `n-quote-safety`; prefer paraphrase and short attributed excerpts.

Documentation

Document	What It Covers
docs/AUDIT.md	Reference-project audit and adopted patterns.
docs/EXAMPLES.md	Detailed example prompts, skill routes, and output shapes.
docs/HARNESS_SUPPORT.md	Claude, Codex, Cursor, Copilot, Gemini, OpenCode, Pi, Kiro, and direct skill install paths.
docs/INSTALL.md	Plugin install, local marketplace, updates, validation, and custom paths.
docs/NAVAL_MEMORY.md	Optional memory setup, privacy contract, folders, schemas, and direct-copy reference layout.
docs/RELEASE.md	Release checklist, validation, smoke, tagging, and GitHub release commands.
docs/skills/	Generated per-skill browsing pages.
docs/SYMLINKS.md	How direct `n-*` skill symlinks work across agent homes.
docs/PLUGIN_REFERENCE.md	Architecture, component map, and plugin-vs-skill tradeoffs.
docs/DEVELOPMENT.md	Regeneration, validation, adding skills, and public repo boundaries.
docs/SOURCE_BOUNDARIES.md	Copyright, attribution, and source-use boundaries.

Local Development

python3 scripts/generate_skill_docs.py
python3 scripts/validate_public.py
python3 scripts/check_coverage.py
python3 scripts/validate_direct_install.py
python3 scripts/smoke_install.py

Build a direct-copy bundle for SKILL.md-only hosts:

npm run export:direct
python3 scripts/validate_direct_install.py --agent-root dist/naval-direct-install

If you have Codex's local validation helpers:

python3 ~/.codex/skills/.system/plugin-creator/scripts/validate_plugin.py .
for d in skills/n-*; do
  python3 ~/.codex/skills/.system/skill-creator/scripts/quick_validate.py "$d"
done

Regenerate generated skill and reference files:

python3 scripts/build_naval_pack.py

Limitations

This is an unofficial skill pack, not an official edition of the book.
It does not include the full book text.
It keeps saved memory opt-in; local .naval/*.local.yaml config files should not be committed.
It should not be used as legal, medical, financial, or mental-health advice.
Exact quotation work should use n-quote-safety and, when needed, verification against an authorized source.

Inspirations

This repository borrows structural ideas from public skill and plugin projects while keeping the Naval content original and paraphrased:

Source Boundaries

This repository contains paraphrased operating knowledge, skill workflows, and coverage maps. It does not include the full text of The Almanack of Naval Ravikant.

The original book is by Eric Jorgenson and is available from Navalmanack.com. This project is not affiliated with Naval Ravikant, Eric Jorgenson, Navalmanack, or any publisher.

License

MIT for the code, generated skill scaffolding, and original documentation in this repository. See LICENSE.

Naval— plugin

Is Naval safe to install?

Unofficial Naval Skills

What The Skills Do

The Loop

Core Workflows

Quick Example

Getting Started

Optional Memory Layer

Harness Support

Components

Skill Areas

Full Skill Inventory

Common Prompts

Install Paths

How It Fits Together

Documentation

Local Development

Limitations

Inspirations

Source Boundaries

License

Audit the pieces. Scan the whole. Decide.