ShaheerKhawaja/build-productionos

build-productionos

ProductionOS smart router — single entry point that routes to the right pipeline based on intent. The ONLY command new users need to know.

Inputs

| Parameter | Values | Default | Description | |-----------|--------|---------|-------------| | intent | string | required | What you want to do. Natural language or keyword. Examples: 'audit this project', 'fix the frontend', 'research authentication', 'review my PR', 'ship it', 'debug the login bug' | | target | string | -- | Target directory, file, or URL (default: current directory) |

/build-productionos — ProductionOS Smart Router

You are the ProductionOS entry point. You receive a natural language intent and route to the right command.

This is the ONLY command a new user needs to know. Everything else is discoverable through this.

Step 0: Preamble

Run the shared ProductionOS preamble (templates/PREAMBLE.md).

Step 0.5: Smart Agent Routing (Production House Layer 1)

Before static intent classification, run the agent index for intelligent agent selection:

bun run "${CLAUDE_PLUGIN_ROOT}/scripts/agent-index.ts" --goal "$ARGUMENTS.intent" 2>/dev/null

Parse the JSON output. If lowConfidence is false, the router has high-confidence agent matches:

• Display: ProductionOS Router → ${agents.length} agents matched (confidence: ${agents[0].confidence})
• Use the matched agents to inform which command to route to (security agents → /production-upgrade, design agents → /designer-upgrade, etc.)
• Pass the agent roster to the routed command so it knows which agents to dispatch

If lowConfidence is true or the script fails, fall through to static keyword matching below.

Stack Detection

Also detect the project stack for tool provisioning:

bun run "${CLAUDE_PLUGIN_ROOT}/scripts/stack-detector.ts" 2>/dev/null

Use the detected stack to inform tool selection in the routed command.

Step 1: Intent Classification (Static Fallback)

Classify $ARGUMENTS.intent into one of these categories:

Category: AUDIT (code quality)

Triggers: "audit", "review", "check", "scan", "grade", "score", "how good is" Route to: /production-upgrade $ARGUMENTS.target

Category: DESIGN (UI/UX)

Triggers: "design", "redesign", "mockup", "UI", "UX", "frontend", "look", "feel", "ugly", "beautiful" Route to: /designer-upgrade $ARGUMENTS.target

Category: UX (user experience)

Triggers: "user story", "journey", "friction", "experience", "onboarding", "flow", "persona" Route to: /ux-genie $ARGUMENTS.target

Category: FIX (targeted improvement)

Triggers: "fix", "improve", "upgrade", "make better", "optimize", "refactor" Route to: /auto-swarm-nth "$ARGUMENTS.intent" --target $ARGUMENTS.target

Category: PLAN (strategic)

Triggers: "plan", "architect", "design system", "strategy", "roadmap", "vision" Route to: /omni-plan-nth $ARGUMENTS.target

Category: BUILD (feature development)

Triggers: "build", "create", "add", "implement", "feature", "new" Route to: /brainstorming $ARGUMENTS.intent → then /writing-plans → then /auto-swarm-nth

Category: DEBUG (fix bug)

Triggers: "debug", "bug", "broken", "error", "crash", "failing", "not working" Route to: /debug $ARGUMENTS.intent

Category: TEST (quality assurance)

Triggers: "test", "QA", "verify", "check", "regression", "e2e" Route to: /qa $ARGUMENTS.target

Category: REVIEW (code review)

Triggers: "review PR", "review code", "review changes", "review diff" Route to: /review

Category: SHIP (deploy)

Triggers: "ship", "deploy", "release", "merge", "push", "PR" Route to: /ship

Category: RESEARCH (deep investigation)

Triggers: "research", "investigate", "explore", "find out", "compare", "what is" Route to: /deep-research $ARGUMENTS.intent

Category: EVAL (self-assessment)

Triggers: "evaluate", "self-eval", "how did I do", "score my work", "rate" Route to: /self-eval session

Category: HELP (guidance)

Triggers: "help", "how do I", "what can you do", "list commands", "getting started" Route to: /productionos-help

Step 2: Confirm Route

Before executing, confirm:

ProductionOS → Detected intent: {CATEGORY}
Routing to: /{command} {arguments}
Target: {target or "current directory"}

If the classification is ambiguous, ask the user to clarify:

I'm not sure if you want:
A) /production-upgrade (audit and fix code quality)
B) /designer-upgrade (audit and fix design/UI)
C) /auto-swarm-nth (parallel targeted fixes)

Which fits better?

Step 3: Execute

Invoke the selected command via the Skill tool or by executing its protocol directly.

Step 4: Post-Route Self-Eval

After the routed command completes, run self-eval:

/self-eval last

Quick Reference (shown on /pos help)

/build-productionos "audit this project"        → Full codebase audit (/production-upgrade)
/build-productionos "fix the frontend"          → UI/UX redesign (/designer-upgrade)
/build-productionos "research auth patterns"    → Deep research (/deep-research)
/build-productionos "review my PR"              → Code review (/review)
/build-productionos "ship it"                   → Ship workflow (/ship)
/build-productionos "debug login bug"           → Systematic debugging (/debug)
/build-productionos "plan the new feature"      → Strategic planning (/omni-plan-nth)
/build-productionos "test everything"           → QA testing (/qa)
/build-productionos "build user profiles"       → Feature development (brainstorm → plan → build)
/build-productionos "how did I do"              → Self-evaluation (/self-eval session)
/build-productionos "help"                      → Usage guide (/productionos-help)

Error Handling

| Scenario | Action | |----------|--------| | No target provided | Ask for clarification with examples | | Target not found | Search for alternatives, suggest closest match | | Agent dispatch fails | Fall back to manual execution, report the error | | Ambiguous input | Present options, ask user to pick | | Execution timeout | Save partial results, report what completed |

Guardrails

1. Do not silently change scope or expand beyond the user request.
2. Prefer concrete outputs and verification over abstract descriptions.
3. Keep scope faithful to the user intent.
4. Preserve existing workflow guardrails and stop conditions.
5. Verify results before concluding. Run self-eval on output quality.