24 Claude Code skill packs target AI code and content quality gaps
Kacper Włodarczyk of Vstorm released two Claude Code skill packs — `production-stack-skills` (10 skills) and `content-skills` (14 skills) — as composable slash commands that audit AI-generated code and content for production readiness and brand consistency.
Score breakdown
Developers can drop these composable, auditable slash commands into any `AGENTS.md`-compatible workflow to get scored, actionable feedback on both production code quality and brand-consistent content — without rewriting their existing agent setup.
- 01Vstorm AI Engineer Kacper Włodarczyk released two skill packs: `production-stack-skills` (10 skills) and `content-skills` (14 skills).
- 02A Claude Code skill is a packaged slash command backed by a `SKILL.md` file, invocable from any `AGENTS.md`-compatible runtime.
- 03`/production check` audits a repo across six categories and returns a 0–100 production-readiness score with a Quick Wins section; internal runs reportedly move scores ~30 points.
Kacper Włodarczyk, AI Engineer at Vstorm (an Applied Agentic AI Engineering Consultancy), has open-sourced two Claude Code skill packs under the MIT license, framing them as a response to two recurring failure patterns observed across 30+ production AI agent deployments: code that passes demos but fails in production, and content that reads as obviously AI-written.
`production-stack-skills` is a 10-skill pack whose flagship command is `/production check`.
`production-stack-skills` is a 10-skill pack whose flagship command is `/production check`. Pointed at a repository, it reads FastAPI routes, Postgres migrations, Dockerfiles, and config files, then returns a headline score (e.g., "Production Readiness: 34/100"), six category scores, a Quick Wins section with point deltas, and an action plan sorted by weighted impact. Internal runs reportedly move scores roughly 30 points from Quick Wins alone. The remaining nine skills — `/production review`, `/production planner`, `/production fastapi`, `/production postgres`, `/production docker`, `/production deploy`, `/production monitoring`, `/production security`, and `/production error-handling` — each target a specific surface area as focused slash commands rather than sub-modes of a monolithic agent.
`content-skills` is a 14-skill pack centered on a `/brand/` directory containing `BRAND.md`, `VOICE.md`, `VISUAL.md`, and voice samples. Running `/content setup` triggers a five-question brand interview that writes this directory once; every subsequent content skill auto-reads it on invocation. The `/content audit` command scores any piece of content 0–100 on voice consistency, anti-slop markers, visual consistency, and brand alignment. The 12 production skills between setup and audit cover strategy, calendar, blog, twitter, linkedin, reddit, hackernews, presentation, infographic, image, video, and repurpose. Both packs install via a single `curl` command that mirrors skills into both `~/.claude/` and `~/.agents/`, making them immediately available across Claude Code, Codex, Amp, and any `AGENTS.md`-compatible runtime without requiring a restart.
Key facts
- 01Vstorm AI Engineer Kacper Włodarczyk released two skill packs: `production-stack-skills` (10 skills) and `content-skills` (14 skills).
- 02A Claude Code skill is a packaged slash command backed by a `SKILL.md` file, invocable from any `AGENTS.md`-compatible runtime.
- 03`/production check` audits a repo across six categories and returns a 0–100 production-readiness score with a Quick Wins section; internal runs reportedly move scores ~30 points.
- 04`content-skills` is built around a `/brand/` directory (`BRAND.md`, `VOICE.md`, `VISUAL.md`) auto-read on every content skill invocation.