Skip to content
Development
Command

/build-app

Build the web client (Next.js, Vite, React) and check env, types and bundle

From plugin
ai-kit
11154 skills15 agents54 commands4 MCP
> /plugin marketplace add solanabr/ai-kit
> /plugin install solana-ai-kit@stbr

Command definition

build-app.md
description: "Build the web client (Next.js, Vite, React) and check env, types and bundle"

Build the web client in $ARGUMENTS (default: the app directory whose `package.json` has a `build` script) for the cluster its program is deployed on. Current client patterns: [frontend.md](../skills/ext/solana-dev/skills/solana-dev/references/frontend.md).

Steps

1. Install with the package manager that owns the lockfile (`pnpm-lock.yaml`, `yarn.lock`, `bun.lock`/`bun.lockb`, `package-lock.json`), in frozen mode (`npm ci`, `pnpm install --frozen-lockfile`). 2. Env: don't read or copy `.env*`, except `.env.example` and `.env.schema`, which hold no values. If `.env.local` is missing, print `cp .env.example .env.local` for the user to run. The firewall covers `.env.local` under the same rules as `.env`, deliberately — it holds the same values. For what is set, use names and presence only (`bash .claude/bin/env-keys.sh .env.local`; if it reports `HELPER_UNAVAILABLE`, as it does at the High tier, ask the user which keys are set rather than guessing); for the program ID, compare `.program-id-devnet` / `.program-id-mainnet` (written by `/deploy`) against what the app logs at runtime, and ask the user to confirm the cluster rather than reading the RPC URL. If the program interface changed, regenerate the client first with `/generate-idl-client`. 3. Type-check with `tsc --noEmit`: Vite builds do not type-check, `next build` does. Run the `lint` script if there is one. 4. Build with `<pm> run build`. Output: `.next/` (Next.js), `dist/` (Vite), `build/` (CRA). 5. Verify: no `*.map` files in the production output unless intended; note the largest JS chunks; preview with `<pm> run start` (Next.js) or `<pm> run preview` (Vite).

Failure fixes

  • `Buffer is not defined` (Vite): add `vite-plugin-node-polyfills` with `nodePolyfills({ include: ['buffer'] })`. web3.js 1.x, the Anchor TS client (`@anchor-lang/core`) and wallet-adapter need it; `@solana/kit` does not.
  • `window is not defined` or hydration errors (Next.js): wallet and web3 code belongs in `'use client'` leaf components.
  • Env var undefined in the browser: client-side vars need the `NEXT_PUBLIC_` or `VITE_` prefix, and a rebuild, because they are inlined at build time.
  • Oversized bundle: lazy-load wallet UI and heavy SDKs; drop `@solana/wallet-adapter-wallets` bundles, since Wallet Standard wallets register themselves; prefer `@solana/kit` over web3.js 1.x for new code.

Guardrails

  • Every `NEXT_PUBLIC_` / `VITE_` value ships in the bundle. Keep RPC URLs that embed an API key (Helius and similar) behind a server route.
  • Never print an env value, including an RPC URL that embeds a key. Report key names and whether they are set.

Output

Framework, package manager, output directory and size, largest chunks, and the env/cluster check result.

Read more
@solanabr
Ships withai-kit

Claude Code, set up to ship on Solana. Not programs alone — the whole lifecycle: validating the idea, building the program and the app around it, shipping it, and the launch and growth work that follows.

1113 skills15 agents4 MCP
Get the whole plugin

Stars on GitHub

Momentum

stars in total
111
Star history of solanabr/solana-ai-kit on GitHub, a chart by star-history.com

Other commands on ai-kit.

add-skill

add-skill

Command

Install a pinned skill extension on demand (/add-skill <id>), or list core packs and extensions

@solanabr@solanabrInstall Command
adv

adv

Command

Short for /adversarial: run every task through proposer and adversary pairs, no triage

@solanabr@solanabrInstall Command
adversarial

adversarial

Command

Put a task's load-bearing claims and key decisions through a proposer and an adversary, with proof

@solanabr@solanabrInstall Command
audit-solana

audit-solana

Command

Audit Solana program code for exploitable bugs and write a findings report

@solanabr@solanabrInstall Command
benchmark

benchmark

Command

Compare per-instruction CU with the stored baseline to catch regressions

@solanabr@solanabrInstall Command

More commands from other plugins.

build
Auto-invoked

build

Command

Implement tasks incrementally — build, test, verify, commit. Add "auto" to run the whole plan in one approved pass.

@addyosmani@addyosmaniInstall Command