refactor(*): remove CLI, convert to pure skills collection

.github/workflows/ci.yml

@@ -1,21 +1,20 @@
-name: CI
+name: Validate Skills
 
 on:
   pull_request:
+    paths:
+      - 'skills/**'
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
 jobs:
-  ci:
+  validate:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-      - uses: oven-sh/setup-bun@v2
-      - run: bun install --frozen-lockfile
-      - run: bun run lint
-      - run: bun test
-      - run: bun run build
-      - run: bun run validate
-      - run: bun run publint
+      - uses: actions/setup-node@v4
+        with:
+          node-version: 22
+      - run: node scripts/validate.js

.gitignore

@@ -1,37 +1,16 @@
-# dependencies (bun install)
+# dependencies
 node_modules
 
-# output
-out
-dist
-*.tgz
-
-# code coverage
-coverage
-*.lcov
-
-# logs
-logs
-_.log
-report.[0-9]_.[0-9]_.[0-9]_.[0-9]_.json
-
-# dotenv environment variable files
+# dotenv
 .env
-.env.development.local
-.env.test.local
-.env.production.local
 .env.local
+.env.*.local
 
 # caches
-.eslintcache
 .cache
-*.tsbuildinfo
 
-# IntelliJ based IDEs
+# IDE
 .idea
 
-# Finder (MacOS) folder config
+# macOS
 .DS_Store
-
-# Cursor
-.cursor/

AGENTS.md

@@ -1,42 +1,52 @@
-# Contributing Agent Instructions
+# AGENTS.md
 
-Instructions for AI agents (Claude Code, Codex, Cursor, OpenCode) contributing to this repo.
+This file provides guidance to AI coding agents working with code in this repository.
 
-## Repository Structure
+## Project Overview
 
-- `skills/` — Agent skill directories, each containing a `SKILL.md` with YAML frontmatter and Markdown instructions
-- `src/` — CLI source (TypeScript, built with Bun)
-  - `src/cli.ts` — Entry point, command routing via Commander
-  - `src/commands/` — Individual command implementations (list, add, remove, info, dev, validate)
-  - `src/agents.ts` — Agent definitions, detection logic, install paths
-  - `src/installer.ts` — Install/uninstall/symlink logic
-  - `src/skills.ts` — Skill discovery and SKILL.md parsing (uses gray-matter)
-  - `src/validator.ts` — SKILL.md validation against the Agent Skills spec
-  - `src/types.ts` — Shared TypeScript types
-- `tests/` — Tests using `bun:test`
+This is a documentation repository containing [Agent Skills](https://agentskills.io/specification) for TxnLab's Algorand ecosystem. Skills extend AI coding agents with specialized knowledge about NFDomains, use-wallet, and Haystack Router. There is no build step or compiled output — the deliverables are markdown files.
 
-## Build, Test, Lint
+## Commands
 
 ```bash
-bun install             # Install dependencies (always use bun, never npm/yarn/pnpm)
-bun run build           # Build CLI to dist/
-bun test                # Run tests
-bun run lint            # Check formatting (Prettier) and linting (ESLint)
-bun run lint:fix        # Auto-fix
-bun run validate        # Validate all SKILL.md files
+# Validate all skills
+node scripts/validate.js
+
+# Validate a specific skill
+node scripts/validate.js <skill-name>
 ```
 
-## Skill Authoring
+There are no build, test, or lint commands. Validation is the only check.
+
+## Architecture
+
+Each skill lives in `skills/<skill-name>/` and follows the Agent Skills specification:
+
+- **`SKILL.md`** — Entry point with YAML frontmatter (`name` and `description` required). The `name` field must match the directory name.
+- **`references/`** — Detailed markdown guides for specific tasks (API usage, framework integration, etc.)
+- **`assets/`** — Optional static assets (none currently used)
+
+The validation script (`scripts/validate.js`) checks frontmatter structure using only Node.js built-ins (no dependencies). It parses YAML frontmatter manually and verifies required fields.
+
+## Current Skills
 
-1. Read `skills/skill-creator/SKILL.md` for the skill creation process
-2. Create a new directory under `skills/<skill-name>/`
-3. Write a `SKILL.md` with valid YAML frontmatter (`name`, `description` required)
-4. Run `bun run validate` to check your skill
-5. Submit a PR — validation runs in CI
+| Skill | Package |
+| --- | --- |
+| `nfd` | `@txnlab/nfd-sdk` — Algorand Name Service (.algo domains) |
+| `use-wallet` | `@txnlab/use-wallet` v4.x — Multi-framework wallet connections |
+| `haystack-router` | `@txnlab/haystack-router` — DEX aggregator and swap routing |
 
-## Testing Conventions
+## Adding a New Skill
 
-- Test files: `tests/<module>.test.ts`
-- Use `bun:test` (describe, test, expect)
-- Tests should be self-contained — create temp dirs, clean up in afterAll
-- Cover both happy paths and edge cases
+1. Create `skills/<skill-name>/SKILL.md` with YAML frontmatter containing `name` (matching directory) and `description`
+2. Add `references/*.md` files for detailed task-specific documentation
+3. Run `node scripts/validate.js` to verify
+4. Use [conventional commits](https://www.conventionalcommits.org/)
+
+## Local Testing
+
+Symlink a skill directory to test with Claude Code directly:
+
+```bash
+ln -s $(pwd)/skills/nfd ~/.claude/skills/nfd
+```

CLAUDE.md

@@ -1,31 +1,3 @@
-# @txnlab/skills
+# CLAUDE.md
 
-Agent skills for TxnLab's Algorand ecosystem. This repo contains curated skills and a CLI for installing them.
-
-## Structure
-
-- `skills/` — Agent skill directories (each with SKILL.md)
-- `src/` — CLI source code
-- `tests/` — Tests (bun:test)
-
-## Commands
-
-```bash
-bun run build           # Build CLI to dist/
-bun run dev             # Run CLI from source (e.g., bun run dev list)
-bun test                # Run tests
-bun run lint            # Check formatting and linting
-bun run lint:fix        # Auto-fix formatting and linting
-bun run validate        # Validate all SKILL.md files
-```
-
-## Conventions
-
-- Always read `skills/skill-creator/SKILL.md` before writing new skills
-- All skills must pass `txnlab-skills validate` before merging
-- Use Prettier for formatting, ESLint for linting
-- Tests use `bun:test`, test files go in `tests/`
-- Commit messages follow conventional commits
-- Bun is the only runtime — do not use npm/yarn/pnpm
-
-See `README.md` for the full CLI reference, supported agents, and contributing guide.
+Strictly follow the rules in ./AGENTS.md