Skills Repository

This repository is a curated catalog of reusable AI skills: compact execution guides that help agents and humans do higher-quality work with less repeated setup. The skills cover software engineering, AI systems, SaaS operations, security, product work, UX, mobile development, documentation workflows, and canonical finance doctrine.

The catalog is designed to be routed by name. A small active surface keeps skill selection reliable, while deeper references stay available without making every topic an active entrypoint.

Why These Skills Matter

Benefit	What it gives you
Faster execution	Reusable workflows reduce repeated prompting and rediscovery.
Better routing	Clear frontmatter, aliases, and parent skills help agents pick the right guidance.
Higher quality	Skills encode checklists, quality gates, anti-patterns, and evidence expectations.
Safer specialization	Finance, security, AI, and platform work keep domain constraints close to implementation guidance.
Portable knowledge	Markdown, YAML, templates, and scripts work across Windows, Ubuntu, and Debian consumers.
Lower catalog noise	Legacy and narrow topics route through aliases instead of competing as duplicate active skills.

Active Catalog

Active skills are SKILL.md files under these roots:

Root	Purpose
`skills/`	Main active catalog for engineering, AI, SaaS, mobile, security, UX, product, and operations.
`doctrine/skills/`	Canonical finance and accounting doctrine for IFRS, controls, close, audit, payroll, inventory, reporting, and finance UX.
`00-meta-initialization/`	SDLC documentation initialization and new-project entrypoints.

Current guardrail baseline:

Metric	Value
Active `SKILL.md` files	169
Target active catalog size	150-170
Guardrail hard cap	200
Duplicate frontmatter names	0
Inactive aliases retained as `ALIAS.md`	47

Run the guardrail report with:

python -X utf8 scripts\skill_catalog_guardrails.py --report-only

Skill Domains

Domain	Examples
AI and agent systems	AI architecture, RAG, evaluations, model gateways, agent runtime, HITL, governance, observability, cost controls, and AI UX.
Software engineering	Architecture, APIs, TypeScript, JavaScript, PHP, Python, Node.js, testing, validation, release engineering, and reliability.
SaaS and product	Multi-tenancy, entitlements, pricing, billing, onboarding, metrics, sales operations, product discovery, and product-led growth.
Security and compliance	Web app audits, code safety, network security, Linux hardening, DPIA work, and AI security controls.
Frontend and UX	React, Next.js, Tailwind, app GUI design, forms, interaction patterns, accessibility, premium UI, and motion.
Mobile	Android, iOS, Kotlin Multiplatform, mobile persistence, mobile UX, platform capabilities, app quality, and release workflows.
Finance doctrine	Accounting engines, finance audits, bank and mobile money reconciliation, close, controls, reporting, IFRS, payroll, inventory, and finance UI patterns.
Documentation and operations	SDLC documentation, project requirements, professional document output, catalog maintenance, skill writing, and update records.

How To Use The Catalog

Start with the active roots above or the overview in docs/overview/README.md.
If an old skill name is mentioned, check docs/skill-routing-index.md or docs/skill-aliases.yml.
Read the selected skill's SKILL.md.
Load only the specific references/, templates/, or scripts/ files the skill tells you to use.
When changing routing, frontmatter, or active skill behavior, update the routing docs and rerun the guardrail report.

Routing And Aliases

The catalog intentionally keeps aliases outside the active skill count. Legacy entrypoints that should no longer compete for routing are retained as ALIAS.md files in their original directories. Their targets are recorded in:

docs/skill-routing-index.md for human-readable policy.
docs/skill-aliases.yml for machine-readable routing.

Finance aliases route to doctrine/skills/ first. Root-level finance skills remain active only when they add implementation or orchestration behavior beyond canonical doctrine.

Repository Map

Path	Role
`docs/`	Overview docs, architecture, routing policy, plans, analysis, and update records.
`scripts/`	Maintenance checks, including active catalog guardrails.
`claude-guides/`	Claude-specific skill creation and invocation guidance.
`book-extractions/`	Curated source notes and long-form reference material.
`blog-posts/`	Draft educational and marketing content.

Maintenance Rules

Do not delete or move skill directories casually.
Deactivate legacy entrypoints by renaming SKILL.md to ALIAS.md and adding a route in docs/skill-aliases.yml.
Keep finance and accounting doctrine canonical under doctrine/skills/.
Keep Markdown files below 500 lines where practical.
Use ASCII unless an existing file requires another character set.
Preserve user edits and inspect the worktree before modifying files.
Record substantive documentation repairs under docs/updates/.

Key Docs

Document	Purpose
`AGENTS.md`	Short working rules for coding agents.
`docs/overview/PROJECT_BRIEF.md`	One-page project brief.
`docs/overview/ARCHITECTURE.md`	Repository structure and ownership boundaries.
`docs/overview/TECH_STACK.md`	Tooling, runtime assumptions, and platform notes.
`docs/plans/INDEX.md`	Planning document index.
`docs/plans/NEXT_FEATURES.md`	Current priorities and next work.

Name	skill-safety-audit
Description	Scan new or updated skills for unsafe or malicious instructions (unknown tools, external installers, credential harvesting) before accepting them into the repository.

skill-safety-audit

SKILL.md