SkillFoxx

Lamina skill

Skills and a CLI with a product graph: roles, states and failures checked before code, the live product after the build

  • 120 ★ on the repository
  • Last commit
  • Apache-2.0
Skill

Extended access

Why this level

  • Installation downloads and runs a CLI binary and a worker from GitHub releases
  • Live-product verification runs the app and works in a browser
All reasons and checks

aryaniyaps/laminaAuthor:aryaniyaps

Install

Manual install

curl -fsSL https://github.com/aryaniyaps/lamina/releases/latest/download/install.sh | sh

First step: installs the lamina CLI from GitHub releases. Then run npx skills add aryaniyaps/lamina --skill '*' -a <agent> -y and lamina setup --agent <agent>.

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Installation downloads and runs a CLI binary and a worker from GitHub releases
  • Live-product verification runs the app and works in a browser

This is third-party code. Review the repository files before installing.

What it does

Lamina keeps product behavior in a local transactional graph for each Git repository and gives the coding agent a contract: what to build, how states and flows work, which failures and edge cases to cover. It creates no app code or styling and imposes no stack. A `lamina` CLI and ten skills go into the project: four workflow entrypoints and six capability skills. After a one-time `/lamina-init` describing the domain and users, ordinary requests such as "add medication schedule editing" run through the chain on their own. Independent walks per persona find missing nodes, permissions and states, `lamina work map` then ties every requirement to implementation and test files, and the live product is verified after the build. `/lamina-design` and `/lamina-verify` are only for a standalone design pass or an audit that writes no code. It suits Codex, Claude Code, Cursor and OpenCode.

Who it is for. Developers building a product with an AI agent who want permissions, states and failures worked out before code, not just the look.

Good fit when

  • A new feature touches several roles and you need permissions and hand-offs checked up front
  • Editing saves with conflicts and you need to describe what a user sees when a save fails midway
  • Before release the built product must be checked per role in the browser and across screen sizes
  • The project is already written and you want to recover its states and flows from the source

Not a fit when

  • You want attractive styling or a page template: the skill is about behavior, not appearance
  • You need an app generator such as v0, Lovable or Bolt: Lamina does not generate application source
  • The change is purely mechanical, such as a rename: a behavior contract adds nothing

Example request

/lamina-init online pharmacy: customers, pharmacists and administrators.

Limitations

The CLI install script downloads a binary and a CocoIndex worker from GitHub Releases and checks them against `SHA256SUMS` before running, which is worth weighing before doing it in a working environment. Node.js and npm are needed only for the `npx skills add` command. After installing you must start a new agent session so the skills appear. The graph is stored locally per repository, and `/lamina-init` creates an unborn `main` branch in a folder without Git but makes no commits. Live-product verification relies on running the app and a browser, so it needs a working environment.

How to disable. Delete the Lamina skills from the agent folder and the lamina binary, then remove the passive rules that lamina setup added. The local graph lives in the project repository and is deleted by hand.

FAQ

How do I install Lamina?

Install the CLI with the `install.sh` script from GitHub releases, then run `npx skills add aryaniyaps/lamina --skill '*' -a <agent> -y` and `lamina setup --agent <codex|claude-code|opencode|cursor>`. Then run `lamina doctor --json` and start a new agent session.

Why not install for every agent with a wildcard?

The README says to install skills only for the active agent: `-a codex`, `-a claude-code` or `-a cursor`. The `-a '*'` variant is discouraged.

How is Lamina different from Impeccable or UI UX Pro Max?

Those skills handle how it looks, Lamina handles how it works: actors, flows, empty and error states, permissions and recovery. You can use them together.

Do I have to call /lamina-design and /lamina-verify?

No. They are advanced overrides for a standalone design pass or audit. In normal work you run `/lamina-init` once and then ask for features in plain language.

What does lamina graph status show?

The state of the repository's active product graph: sources, versions and failures. The project reference describes the graph and all commands in full.

Discussion

No questions or reviews yet.

Skills

Lattice: composable skills for AI assistants covering requirements, design before code, standards-based work and review

PluginMinimal access211

lencx skills

lencx-skills

Two skills for Claude Code and Codex: keel works through architecture questions, coding-protocol keeps code edits inside the task

SkillMinimal access197

Standalone skills for the whole development lifecycle: product, architecture, planning, implementation, QA, delivery and operations

SkillBroad access594

Ten software lifecycle skills for Claude Code: requirements, design, tests, adversarial review and release checks with hooks.

PluginExtended access66

Garry Tan's skill suite that turns Claude Code into a virtual team: product, architecture, design, review, QA and shipping

SkillBroad access141.4K

Twelve standalone skills for development: an agent loop with planning and review, pull request handling, TDD and email triage.

SkillExtended access56

README in short

The README moves from a quick start and one-time init to the loop: design, independent persona walks, build and live-product verification. It then explains the local graph, stack compatibility, a hotel-booking demo and comparisons with Impeccable, BMAD, Spec Kit, v0 and Figma. A command table and development notes close it.

README badge

Original SKILL.md

---
name: lamina
description: "Route ordinary product implementation and explicit Lamina lifecycle requests. Use passively after initialization for changes to product behavior, flows, permissions, states, failures, or UI; skip purely mechanical maintenance."
---

# /lamina

Use this as the compact Lamina router. The public catalog has four workflow

Report a problem