mnguyenht

bld-orchestrator-fable

of: Boss mode over Claude subagents. Plan a brain-dump, fan out bld-executor workers in parallel, judge their work. Use for a messy multi-item to-do list or several independent pieces to build at once.

bld-orchestrator-opus

oo: The /bld-orchestrator-fable loop with Opus as the boss. Use when the user wants the orchestrator run on Opus.

bld-find-21st

f2: Find a ready-made shadcn/Tailwind component on 21st.dev, shortlist options, install the pick. Use for "find a component", "get me a navbar/hero/card", or premade UI instead of building from scratch.

bld-find-spline

fs: Find a ready-made 3D scene on Spline and wire its embed into the app. Use for "find a 3D scene", "add a Spline", "3D hero". Heavy. Only when it genuinely elevates the page.

bld-mcp-settings

Switch BLD's MCP servers between on-demand and always-on by editing .mcp.json. Use for "keep my MCPs running", "always-on MCP", "stop spawning MCPs every time", or when an MCP server is not connected.

bld-optimize-app

oa: Run Lighthouse on the live app (performance, accessibility, SEO), triage, fix what's real. Use for "run Lighthouse", "check my PageSpeed", "why is my site slow", "audit the live site", or before a public launch. Source code is /bld-optimize-react.

bld-optimize-react

or: Scan React/Next source for real defects (bugs, hooks, re-renders, a11y) with react-doctor and eslint, triage and fix. Use for "scan this app", "audit this app", "check for re-renders", or before shipping a screen. Live site is /bld-optimize-app.

bld-optimize-security

os: Full static security audit of an app across 13 layers, front end to recovery. Reads code and config, never attacks. Use for "secure this app", "full security audit", "harden this", or before shipping something serious.

bld-optimize-seo-indexing

oi: Get an app found on Google. Audit the live site, add sitemap, robots, canonical and structured data, then list the account steps only the user can do. Use for "get this on Google", "index my site", "why isn't my site showing up", "SEO this".

Data & Documents Listed

bld-professional-settings

Switch BLD command names between friendly and pro mode (which word comes first). Use for "pro mode", "shorter commands", "go back to the long names", or which naming mode is on.

bld-quiz

qz: Post-sprint learning check. Quiz the user on what was just built, sized to the sprint; tiny changes get a walkthrough instead. Use for "quiz me", "test me on that", or when a sprint wraps.

bld-runtime-activate-mcps

rm: Spawn an MCP server (jcodemunch, context-mode, shadcn) for one burst of code search or registry lookup, then shut it down. Use for "hard optimize", "token-optimize this search", or a big multi-file sprint.

bld-runtime-agents

ra: Boss mode over an external coding agent. Spec small pieces, have Codex (Gemini fallback) build them, review each against the spec. Use for "delegate this", "have Codex build X", or bulky repetitive work.

bld-runtime-tokens

rt: Check how much of the Claude token window is left and say what to do. Continue, batch, delegate, or hand off. Use for "how many tokens left", "am I about to run out", "check my usage", before a long sprint, or unprompted when one runs long.

bld-settings-block-image-generation

confi: The one switch for BLD's AI image-generation block, which is on by default. Use it to turn the block off, put it back, or check where it stands. Also for "allow image generation", "no AI images", "is the image hook on".

bld-setup

st: Install BLD or add more of it. Checks prerequisites, lists every tool with its source, installs only what the user picks, resumes if interrupted. Use for "set up BLD", "install the BLD toolkit", "what else can I install".

bld-sprint-init

si: Take an app idea to a good-looking, working, deployed base. Design direction, Vite + React + shadcn scaffold, core screens, deploy. Use for "start a new app", "build me an app", "spin up a new project". Polish is /bld-sprint-refine.

bld-sprint-planning

sp: Decide what to build before building. Interview the idea, test its core assumption, cut to a shippable v1, write the planning files. Use for "I have an idea", "help me plan this", "plan the next feature", "what should I build". No app code.

bld-sprint-refine

sr: Polish a working app base screen by screen. Design craft, motion, optional 3D and premade components. Use for "polish this app", "make it feel great", "bring it to life", once a screen is basically done.

bld-util-copywriting

uc: Write or fix app copy (hero, CTAs, features, empty states, pricing) with no invented proof and no AI-sounding phrasing. Use for "write the copy", "fix this headline", "this sounds like AI", or placeholder text.

bld-util-customize-component

us: Build a temporary slider panel so the user tunes an effect or animation by hand, then commit the values. Use for "let me tune this", "give me sliders", "control panel", or when a visual tweak has missed twice by prompt.

DevOps & Infrastructure Listed

bld-util-deploy

ud: Deploy an app to a private GitHub repo + Vercel with auto-deploy on push. Use for "deploy this", "put this online", "get me a share link".

Data & Documents Listed

bld-util-documentation

um: Write a docs subpage inside the app, in Simplified Technical English, covering only what the code actually does. Use for "write the docs", "document this app", "add a docs page", "user guide", "help page".

bld-util-handoff

uh: Write handoff.md (goal, state, changes, failed attempts, next steps) so work continues cleanly in a fresh session. Use for "hand off", "write a handoff", "save state and start fresh".

bld-settings-mcp

confm: Switch BLD's MCP servers between on-demand and always-on by editing .mcp.json. Use for "keep my MCPs running", "always-on MCP", "stop spawning MCPs every time", or when an MCP server is not connected.

Data & Documents Listed

bld-settings-professional

confp: Switch BLD command names between friendly and pro mode (which word comes first). Use for "pro mode", "shorter commands", "go back to the long names", or which naming mode is on.