Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Stack

ASecurity

Build the canonical AgentikOS app stack — Next.js + Convex + Clerk + Stax, with Stripe as an opt-in flag — from scratch or from a Blueprint OS blueprint folder. Pulls the live Stax checkout before every scaffold (R-BLUEPRINT-STACK), vendors the panel engine, wires Convex auth to Clerk, and (with --stripe only) stubs subscriptions with an explicit list of the IDs the operator must provide. Use when the user says "/stack", "/omg-stack", "new app", "scaffold the stack", "build the app", "start a...

2 stars
0 votes
0 copies
0 views
Added 10/2/2026
ai-agentstypescriptgoshellbashreactnextjsgitfullstack

Works with

cli

Security Analysis

A92/100
mediumInstalls packages at runtime which could introduce malicious dependencies

Pro scans all 6 files and shows the line behind each finding

Scanned 10/2/2026

$npx -y skills add agentik-os/claude-code-skills --skill stack --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Stack?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Stack
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/agentik-os-stack/badge)](https://www.skillsdirectory.com/skills/agentik-os-stack)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: stack
description: >
  Build the canonical AgentikOS app stack — Next.js + Convex + Clerk + Stax, with Stripe
  as an opt-in flag — from scratch or from a Blueprint OS blueprint folder. Pulls the
  live Stax checkout before every scaffold (R-BLUEPRINT-STACK), vendors the panel engine,
  wires Convex auth to Clerk, and (with --stripe only) stubs subscriptions with an
  explicit list of the IDs the operator must provide. Use when the user says "/stack", "/omg-stack", "new app",
  "scaffold the stack", "build the app", "start a new OS", "nouveau projet Next Convex",
  "monte la stack", "construis l'app", "build from the blueprint", or names any subset of
  Next.js / Convex / Clerk / Stripe / Stax as the thing to set up. This is the BUILD half
  of the Blueprint OS chain: /blueprint-os designs, /stack builds.
triggers: ["stack","omg-stack","new app","scaffold the stack","build the app","start a new OS","build from blueprint","nouveau projet","monte la stack","construis l'app","next convex clerk","stax app"]
allowed-tools: ["Bash","Read","Write","Edit","Grep","Glob"]
domain: fullstack
read_only: false
argument-hint: "<app-name> [--blueprint <path-to-blueprint-folder>] [--dir <parent-dir>]"
source: OmegaOS — pairs with skills/blueprint-os and skills/stax
---

> **OmegaOS skill — the build half of the Blueprint OS chain.**
> `/blueprint-os` designs an OS. `/stack` turns that design into a running app on the
> canonical stack. Triggers `/stack` and `/omg-stack`.

## The stack, non negotiable (R-BLUEPRINT-STACK)

| Layer | Techno | Role |
|---|---|---|
| **Shell / UI** | **Stax** | The panel grammar. Every OS wears the same navigation |
| Front | Next.js (App Router) | The app |
| Data + realtime | Convex | The single model. Reactive by default |
| Auth | Clerk | Identity, organizations, roles |
| AI | Claude, called from a **Convex action** | Never from the client |

If a design suggests something else, the design changes, not the stack.

**Stripe is opt-in, not part of the default.** Pass `--stripe` when the app actually
charges someone. Most OS products are built and used long before they are sold, and a
billing surface nobody calls is dead code that still demands keys, a webhook endpoint,
and a dashboard setup. Scaffold it when it earns its place.

## Stax is pulled every single time

Before any scaffold, `stack-new.sh` runs the Stax sync — fast-forward only, never
clobbering local commits. The app is always vendored from the **current** `main` of
`github.com/agentik-os/stax`. The commit that was vendored is written into
`stax.lock.json` at the app root, so a build is always traceable to a Stax revision.

This is the rule, not an optimization: an app scaffolded from a stale Stax drifts from
the rest of the family, and the drift is invisible until the panel grammar disagrees.

---

## How to run it

```bash
bash ~/.omega/skills/stack/scripts/stack-new.sh <app-name> [--blueprint <dir>] [--dir <parent>] [--stripe]
```

| Flag | Effect |
|---|---|
| `--blueprint <dir>` | Read a Blueprint OS folder: phase `09-data/schema.ts` becomes the Convex schema, phase `10-stax/panneaux.md` drives the panel registry |
| `--dir <parent>` | Where the app is created. Defaults to `~/Station/SideBusiness` |
| `--stripe` | **Opt-in billing.** Adds the checkout route, the webhook route, `lib/stripe.ts`, the `stripe` dependency, and the Stripe env block |
| `--no-install` | Scaffold the files, skip `npm install` (offline / inspection) |

The script is **idempotent on a fresh dir and refuses to overwrite an existing app** —
rerunning it on a live project is a destructive act, so it stops instead (R-DESTRUCT).

---

## What the script does, in order

1. **Sync Stax** — `stax-sync.sh`, ff-only. Records the commit.
2. **`create-next-app`** — TypeScript, App Router, Tailwind, `src/`.
3. **Vendor Stax** — calls the existing `stax-scaffold.sh` (never reimplemented here):
   `panels-core.ts`, `panels-react.tsx`, `tokens.css`, `stax-ui.css` into `src/stax/`.
4. **Convex** — `convex/schema.ts` (from the blueprint if given, else the canonical
   pattern), `convex/auth.config.ts` wired to Clerk, an example query and mutation.
5. **Clerk** — middleware, provider, and the Convex `ConvexProviderWithClerk` bridge.
6. **Stripe — only with `--stripe`** — checkout route, webhook route, and `lib/stripe.ts`.
   The **price IDs are left as named placeholders** because only the operator can create
   them. Without the flag, none of this is written and `stripe` is not installed.
7. **`.env.example`** — every key, with who provides it and where to get it.
8. **`NEEDS-OPERATOR.md`** — the exact list of what a human must do before the app runs.

Then it prints the list. It never invents a key, never fakes a Stripe ID, and never
claims the app runs before `npm run dev` has actually been executed.

---

## The Stripe IDs — read this before promising anything

*Only relevant when the app was scaffolded with `--stripe`.*

Stripe is the one layer that **cannot** be fully scaffolded. Products and prices live in
the operator's Stripe account, and their IDs (`price_…`, `prod_…`) do not exist until a
human creates them. The scaffold therefore writes placeholders and lists them.

Never guess a `price_…`. A wrong price ID fails at checkout, in production, on a real
customer. See `references/stripe.md` for the full setup path and the webhook events
that actually matter.

---

## Building from a blueprint

When `--blueprint` points at a Blueprint OS folder, the chain is:

| Blueprint phase | Becomes |
|---|---|
| `09-data/schema.ts` | `convex/schema.ts`, verbatim if valid |
| `10-stax/panneaux.md` | The panel registry and the inspector stubs |
| `06-features/features.md` | The `NEEDS-BUILD.md` checklist, layer by layer |
| `08-ia/system-prompts.md` | `convex/ai/` action stubs, one per agent |

**Gate, enforced mechanically.** `stack-new.sh` runs `blueprint-check.sh --gates-only`
before it scaffolds anything, and **refuses** on an unfranchised gate. Building from a
blueprint that failed the parity gate produces a demo whose missing socle is discovered
at delivery, which is exactly the failure mode phase 5 exists to prevent.

`--force-gates` overrides it. Use it knowing what it costs, and it says so out loud.

Run the full check yourself when you want more than the gates:

```bash
bash ~/.omega/skills/blueprint-os/scripts/blueprint-check.sh <blueprint-dir>
```

---

## Rules of conduct

**Pull Stax first, always.** No exception, no "it was synced this morning".

**Never fabricate an ID or a key.** Placeholder plus a line in `NEEDS-OPERATOR.md`.

**Never start a dev server to "verify" (R-TEST)** — except here, where the code is
brand-new and not yet deployed. That is the documented exception: one `npm run dev`,
one HTTP probe, then kill it. Real runtime evidence, not a green install log (L1).

**Secrets never enter the repo (R-ENV / L0).** `.env.local` is gitignored; the
`.env.example` carries names only. Keys live in `~/.omega/secrets/<app>.env`.

**The panel grammar is not optional.** No pages, no modals, no tabs. If the app grows a
second navigation mechanic, it stopped being an AgentikOS app.

---

## Reference files

| File | When |
|---|---|
| `references/convex.md` | Schema, indexes, auth, and calling Claude from an action |
| `references/clerk.md` | The Clerk plus Convex bridge, and the JWT template that trips everyone |
| `references/stripe.md` | Products, prices, webhooks, and the IDs the operator must create |
| `references/env.md` | Every environment variable, who provides it, where it lives |

Attribution

agentik-osagentik-os
View sourceSee grades on GitHubMore from agentik-os →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698461 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →