README.md
July 19, 2026 Β· View on GitHub
The official best-practices package for grimoire β curated, peer-reviewed expert practices across 27 domains. Declare the practices your project follows. Enforce them in CI with grimoire check.
π grimoire-core π
The official package. Declare the practices you need, install them in one command,
enforce compliance with grimoire check β across every AI agent, every project.
Why | Install | Quick Start | Agents | Domains | Standard | Contributing | Changelog | License
π€ Why grimoire-core?
Your AI knows everything β Grimoire makes it practice it.
Books gave everyone knowledge. Google gave everyone access. AI gave everyone comprehension. None of them gave everyone practice. Grimoire does.
The model knows SOLID, DDD, Google SRE, and the ABA Model Rules. Without explicit guidance, it enforces none of them. This library is the enforcement layer β skills: named, citable, executable units of expert practice. One concept. One source. One set of steps.
- π You don't know what you don't know. Grimoire surfaces the governing standard you didn't know applied.
- π€ Knowing is not enough. Grimoire enforces. Specific steps, verifiable criteria, repeatable results. Not summaries β verdicts.
- π The world's best practices belong to everyone. McKinsey charges $1M. Senior lawyers bill $800/hr. The practices they follow are not proprietary. Grimoire makes them free β as executable steps, cited and verified.
- π§± Every profession. Every AI. 1000+ skills across 27 domains. Works with Claude, Copilot, Codex, Cursor, Gemini, OpenCode, OpenClaw, and Antigravity.
- π¬ Not just distribution β enforcement. Skills registries distribute files. grimoire-core provides the practice rules
grimoire checkvalidates your project against. 1000+ skills are both instructions for your AI and compliance criteria for your codebase.
| Repo | Role | Unique value |
|---|---|---|
| jeffreytse/grimoire | CLI β package manager + BPDD linter | grimoire check enforces declared practices |
| jeffreytse/grimoire-core (this repo) | Official package β 27 domains, STANDARD.md quality gate | Semantic compliance rules, not just skills |
| Any git repo | Community / private package | grimoire installs and manages it β standard encouraged, not required |
If you've spent 10,000 hours mastering something, your practice belongs here.
π οΈ Install
macOS / Linux:
# 1. Install the grimoire CLI
curl -fsSL https://raw.githubusercontent.com/jeffreytse/grimoire/main/scripts/install.sh | bash
# 2. Add + install grimoire-core in one step
grimoire init # creates grimoire.toml, auto-detects AI agents
grimoire install jeffreytse/grimoire-core # adds to grimoire.toml + links to all AI agents
Go:
go install github.com/jeffreytse/grimoire@latest
grimoire init && grimoire install jeffreytse/grimoire-core
grimoire.toml β declare grimoire-core as a dependency:
# grimoire.toml β commit this to your repo
[package]
name = "my-project"
[dependencies]
"jeffreytse/grimoire-core" = "*" # always latest
# "jeffreytse/grimoire-core" = "~1.0" # pin to 1.x
# "jeffreytse/grimoire-core" = "1.2.3" # exact pin
Native plugin install (Claude Code):
/plugin marketplace add jeffreytse/grimoire-core
/plugin install grimoire@grimoire-core # all domains
/plugin install grimoire-engineering@grimoire-core # one domain
π€ Agent Support
| Agent | Plugin install | Script install |
|---|---|---|
| Claude Code | /plugin marketplace add jeffreytse/grimoire-core then /plugin install grimoire@grimoire-core | grimoire install --target claude |
| GitHub Copilot CLI | copilot plugin marketplace add jeffreytse/grimoire-core then copilot plugin install grimoire@grimoire-core | grimoire install --target all |
| Gemini CLI | gemini extensions install https://github.com/jeffreytse/grimoire-core | grimoire install --target gemini |
| OpenCode | See .opencode/INSTALL.md | grimoire install --target opencode |
| Codex CLI | AGENTS.md auto-loaded | grimoire install --target codex |
| Cursor | AGENTS.md context injection | grimoire install --target cursor |
| OpenClaw | See .openclaw/INSTALL.md | grimoire install --target openclaw |
| Antigravity CLI (agy) | See .agy/INSTALL.md | grimoire install --target antigravity |
π Quick Start
Describe any problem in plain language β grimoire routes to the right skill:
User: I need to raise a Series A but don't know how to pitch investors.
Claude: Situation matches: write-value-proposition + design-go-to-market + apply-pyramid-principle
β Start with your value prop. /write-value-proposition
Or invoke a skill directly:
/suggest-best-practice # describe any problem β auto-routes to the right skill
/review-pull-request # engineering code review
/calculate-fire-number # how much do I need to retire?
/review-saas-contract # flag the 3 highest-risk clauses
/design-sleep-protocol # evidence-based sleep improvement
/negotiate-salary # structure your compensation negotiation
Common paths:
| You want to⦠| Use |
|---|---|
| Know exactly which skill you need | /skill-name directly |
| Have a problem, unsure which skill | /suggest-best-practice |
| Already have a plan, want gaps checked | /review-best-practice-fit |
| Need 2+ practices coordinated in sequence | /plan-best-practice-solution |
| Complex problem β sub-problems emerge during execution | /apply-best-practice-tree |
| Don't know what practices exist for a topic | /discover-best-practices |
| About to start a task β want to catch gaps first | /start-best-practice |
π¦ Using as a Package
grimoire-core is the official package of the grimoire skills manager β one package among many, just the curated one. Declare it in grimoire.toml:
[dependencies]
"jeffreytse/grimoire-core" = "*" # all 1000+ skills
"jeffreytse/grimoire-core:engineering" = "*" # engineering domain only
"jeffreytse/grimoire-core:health" = "*" # health domain only
"myorg/internal-skills" = "*" # your private package
Versioning follows semver. Pin with "~1.0" (patch updates) or "1.2.3" (exact). Use "*" to track the latest verified release.
The :path suffix filters which skills are linked. No : = all skills from that package. Paths are Standard Glob patterns (doublestar) matched against the skill's domain path. ** matches any directory depth; * matches within a single segment; ?, {a,b}, and [...] are also supported. Examples: engineering/** (all engineering subdomains), health/sleep/** (one subdomain), **/development (development subdomain in any domain).
Installing grimoire-core gives you both: 1000+ skills your AI can invoke, and the practice rules grimoire check validates your project against. One package, two modes.
πΊοΈ Domains
| Domain | Sub-domains |
|---|---|
| meta | suggest-best-practice, plan-best-practice-solution, discover-best-practices, learn-best-practice, learn-grimoire, teach-best-practice, explain-best-practice, write-best-practice-skill, review-best-practice-skill, β¦ |
| engineering | development, frontend, architecture, testing, reliability, devops, cloud, networking, security, data, ai, hardware, mobile, performance, project-management, product, documentation |
| business | strategy, operations, leadership, entrepreneurship, hr |
| science | biology, physics, chemistry, mathematics, earth-science, astronomy |
| health | fitness, nutrition, mental-health, sleep, medicine |
| writing | creative, technical, copywriting, academic, journalism |
| design | ui-ux, graphic, branding, motion, product |
| marketing | seo, content, social-media, paid-ads, growth, analytics |
| finance | personal-finance, investing, accounting, real-estate, corporate |
| law | contracts, ip, employment, privacy, corporate |
| education | curriculum, teaching, e-learning, assessment, learning-science |
| film | cinematography, directing, editing, screenwriting, production |
| photography | composition, lighting, editing, genres |
| music | composition, production, mixing, theory, performance |
| cooking | techniques, baking, flavor, nutrition, world-cuisine |
| language | learning, linguistics, translation, communication |
| art | drawing, painting, digital-art, illustration, color-theory |
| sports | training, coaching, nutrition, tactics, recovery |
| productivity | time-management, habits, focus, goals, tools |
| travel | planning, budgeting, cultural, adventure |
| psychology | cognitive, behavioral, social, clinical, positive |
| home | renovation, interior-design, gardening, organization, smart-home |
| environment | sustainability, ecology, climate, energy, policy |
| pets | dogs, cats, training, nutrition, health |
| fashion | styling, wardrobe, design, sustainability, accessories |
| parenting | infant, toddler, school-age, teen |
| automotive | maintenance, troubleshooting, buying, modifications, ev |
π Featured Skills
| Skill | Domain | Source methodology | Verified |
|---|---|---|---|
review-saas-contract | law/contracts | ABA model SaaS agreements | β |
calculate-fire-number | finance/personal-finance | Bengen (1994) / Trinity Study | β |
negotiate-salary | finance/personal-finance | Fisher & Ury "Getting to Yes" | β |
design-sleep-protocol | health/sleep | Walker "Why We Sleep" / AASM | β |
apply-mise-en-place | cooking/techniques | Culinary Institute of America | β |
apply-five-whys | engineering/reliability | Toyota Production System | β |
review-pull-request | engineering/development | Google Engineering Practices | β |
design-training-program | sports/training | NSCA / Bompa "Periodization" | β |
Law β audit a SaaS contract before signing:
User: /review-saas-contract
Claude: Red flags found.
π΄ CRITICAL Liability cap is \$500 β your contract value is \$240,000
π΄ CRITICAL No data processing agreement β GDPR Art.28 requires one
before processing any EU customer data
π‘ WARNING Auto-renewal clause: 90-day cancellation window in Β§14.3
π‘ WARNING IP ownership of custom integrations is unassigned
β PASS SLA 99.9% with defined service credits
Do not sign. Renegotiate: liability cap, DPA, IP clause.
Source: ABA model SaaS agreements
Finance β calculate how much you need to retire:
User: /calculate-fire-number
Claude: Applying Trinity Study model (Bengen 1994, 95% 30-year success rate)
Annual expenses: \$60,000
Safe withdrawal rate: 4.0%
βββββββββββββββββββββββββββββ
FIRE number: \$1,500,000
At \$4,000/month invested + 7% real return β retire in 20.4 years
Cut annual spend by \$12,000 β retire in 16.8 years
Next: /design-portfolio-allocation
Sports β design a training program:
User: /design-training-program
Claude: Hypertrophy block β intermediate, 4 days/week
UPPER (Mon/Thu) Bench 4Γ8 Β· Row 4Γ8 Β· OHP 3Γ10 Β· Face pull 3Γ15
LOWER (Tue/Fri) Squat 4Γ6 Β· RDL 3Γ8 Β· Leg press 3Γ12
Progressive overload: +2.5 kg when all reps completed clean
Week 5: deload β 40% volume, same intensity
Source: NSCA CSCS + Israetel volume landmarks (2019)
β Browse all skills by domain
π The Grimoire Skill Standard
grimoire maintains an open standard for AI agent skill quality β freely adoptable by any skill library.
Every skill must pass review-best-practice-skill before merge:
| Criterion | Requirement | Rejection example |
|---|---|---|
| Adopted by | Named organizations or institutions | "Many top companies" |
| Impact | Cited study or % number | "Significantly improves quality" |
| Steps | Immediately executable | Abstract theory or advice |
| Scope | One concept per skill | "Nutrition and training program" |
| Source | External institution or standard body | Internal opinion |
Each skill lives at skills/<domain>/<subdomain>/skills/<skill-name>/SKILL.md:
---
name: skill-name
description: Use when <triggering conditions>
source: Author/Org, "Title", Year
tags: [problem-keyword, tool, role, outcome]
---
# Skill Name
One-sentence purpose.
## Why This Is Best Practice
**Adopted by:** ...
**Impact:** ...
## Steps
...
β Read the full standard Β· Adopt this standard
π€ Contributing
Your first skill in ~30 minutes:
- Pick a practice you've used at the highest level in your field
- Run
/write-best-practice-skillβ it guides you through the format step by step - Open a PR β
/review-best-practice-skillruns automatically and flags any gaps - Merge after review passes
This workflow applies to contributions to this repository (the official package). Skills in your own package have no such requirement β see docs/user-package.md.
| Task | Skill |
|---|---|
| Write a new skill | write-best-practice-skill |
| Review a skill PR | review-best-practice-skill |
| Fix review findings | revise-best-practice-skill |
| Add a new domain | design-best-practice-domain |
| Audit a domain's health | audit-best-practice-domain |
| Retire an outdated skill | deprecate-best-practice-skill |
See CONTRIBUTING.md Β· GOVERNANCE.md Β· STANDARD.md
π οΈ Build Your Own Package
You don't have to contribute here. Any git repo is a grimoire package β no spec required.
Three tiers:
| Tier | Requirement | What you get |
|---|---|---|
| Unofficial | Any file layout | Works with grimoire install myrepo/my-skills |
| Compatible | SKILL.md frontmatter (name, description, tags) | Skill metadata, grimoire list, search |
| Standard | Passes STANDARD.md review | PR welcome here, or publish as a standalone package |
Quickstart:
# Create a new skills package
mkdir my-skills && cd my-skills
git init
mkdir -p skills/engineering/development/skills/my-skill
cat > skills/engineering/development/skills/my-skill/SKILL.md << 'EOF'
---
name: my-skill
description: Use when ...
tags: [engineering, development]
---
# My Skill
...
EOF
# Push to GitHub, then add + install in any project:
grimoire install yourusername/my-skills # all skills
grimoire install "yourusername/my-skills:engineering/**" # one domain only
grimoire update # keep it up to date
Publish as grimoire-<name> (e.g. grimoire-cooking, grimoire-fintech) to make it discoverable by others.
β Full package authoring guide and ecosystem overview in the grimoire README.
π§ Maintainer
Update plugin manifest before each release (keeps Claude Code marketplace in sync with new skills):
make update-plugin-manifest # auto-generates .claude-plugin/plugin.json from directory structure
make check-manifest # CI gate: fails if manifest is out of sync
π License
This project is licensed under the MIT license Β© Jeffrey Tse