| name | blueprint |
| description | Canonical Roblox Studio consensus-planning workflow for creator architecture, scope, and implementation guidance |
| surface-class | canonical |
| domain | roblox-studio |
| audience | creator |
| artifact-type | skill |
Blueprint
Blueprint is the canonical creator planning workflow in RCS.
Use $blueprint to turn a creator brief into an execution-ready Roblox Studio plan with the right vocabulary, service boundaries, data flow, and verification path.
Treat creator-docs as canonical when gameplay, UX, monetization, psychology, or Roblox-specific structure already exists in this repo.
Blueprint is the canonical creator-facing equivalent of consensus planning. It drives iterative planning with Planner, Architect, and Critic agents until the plan is explicit, testable, and ready for either $forge or $team.
Usage
$blueprint "task description"
$blueprint --interactive "task description"
Flags
--interactive: Enables user prompts at key decision points using the structured question UI.
--deliberate: Forces deliberate mode for high-risk work, including pre-mortem scenarios and expanded test planning.
Purpose
Blueprint exists to stop planning drift before implementation:
- convert vague feature ideas into Roblox-native architecture
- name the real engine concepts and services involved
- define server/client ownership before a single script changes
- keep the repo from defaulting to generic enterprise abstractions that do not fit a Roblox experience
What Blueprint Must Produce
- the concrete creator outcome
- the gameplay/system scope
- the exact Roblox services and object model touchpoints
- server/client/shared ownership boundaries
- module/script placement guidance
- data shape guidance using Luau/Roblox terms
- verification steps tied to the changed gameplay or tool flow
Roblox-Native Planning Rules
- Prefer engine truth over generic software jargon.
- Talk in terms of:
- Instances
- services
- remotes
- attributes
- tags
- replicated state
- server authority
- player lifecycle
- UI trees
- ModuleScripts
- typed Luau
- If a plan introduces patterns that feel imported from enterprise backend work, justify them explicitly.
- If a system can be expressed more directly through normal Roblox structures, choose that simpler path.
GPT-5.5 Guidance Alignment
Use the shared workflow guidance pattern: outcome-first framing, concise visible updates for multi-step planning, local overrides for the active workflow branch, evidence-backed planning and validation expectations, explicit stop rules, right-sized implementation/PRD shape, and automatic continuation for safe reversible steps. Ask only for material, destructive, credentialed, external-production, or preference-dependent branches.
Consensus Workflow
Blueprint invokes the consensus-planning lane:
$plan --consensus <arguments>
$plan --consensus --interactive <arguments>
The consensus workflow:
- Planner creates an adaptive plan and a compact BLUEPRINT-DR summary before review. Do not default to exactly five steps; right-size the plan to the real scope:
- Principles (3-5)
- Decision Drivers (top 3)
- Viable Options (>=2) with bounded pros/cons
- If only one viable option remains, explicit invalidation rationale for alternatives
- Deliberate mode only: pre-mortem (3 scenarios) + expanded test plan (unit/integration/e2e/observability)
- User feedback
(--interactive only): Present the draft plan plus the Principles / Drivers / Options summary before review.
- Architect reviews for architectural soundness and must provide the strongest steelman antithesis, at least one real tradeoff tension, and, when possible, synthesis.
- Critic evaluates against quality criteria. Run only after Architect completes.
- Re-review loop (max 5 iterations): Any non-
APPROVE Critic verdict must run the full closed loop of revise -> Architect -> Critic until approval or iteration cap.
- On Critic approval
(--interactive only): Present the final plan with approval options:
- Approve and execute via canonical
$forge
- Approve and implement via
$team
- Request changes
- Reject
- Final plan output MUST include:
- ADR (Decision, Drivers, Alternatives considered, Why chosen, Consequences, Follow-ups)
- available-agent-types roster
- follow-up staffing guidance for both
$forge and $team
- suggested reasoning levels by lane
- explicit
rcs team / $team launch hints
- a concrete team verification path
- On approval: MUST hand off to
$forge or $team. Do not implement directly from the planning agent.
Important: Steps 3 and 4 MUST run sequentially. Await completion before step 4. Do NOT run Architect and Critic in parallel.
Pre-context Intake
Before consensus planning or execution handoff, ensure a grounded context snapshot exists:
- Derive a task slug from the request.
- Reuse the latest relevant snapshot in
.rcs/context/{slug}-*.md when available.
- If none exists, create
.rcs/context/{slug}-{timestamp}.md (UTC YYYYMMDDTHHMMSSZ) with:
- task statement
- desired outcome
- known facts/evidence
- constraints
- unknowns/open questions
- likely codebase touchpoints
- If ambiguity remains high, gather brownfield facts first. When session guidance enables
USE_RCS_EXPLORE_CMD, prefer rcs explore for simple read-only repository lookups with narrow, concrete prompts; otherwise use the richer normal explore path. Then run $deep-interview --quick <task> before continuing.
- If the plan depends on official docs, version-aware framework guidance, best practices, or external dependency behavior, auto-delegate
researcher before finalizing the planning handoff.
Do not hand off to execution modes until this intake is complete. If urgency forces progress, explicitly document the risk tradeoffs.
Pre-Execution Gate
Why the Gate Exists
Execution modes ($forge, autopilot, team, ultrawork) are too expensive to launch against a vague request. Blueprint intercepts underspecified execution asks and redirects them into explicit planning first, so code execution starts from a bounded plan instead of a guess.
This ensures:
- explicit scope
- test specification
- planner/architect/critic consensus
- no wasted execution cycles
Good vs Bad Prompts
Passes the gate:
forge fix the null check in src/hooks/bridge.ts:326
autopilot implement issue #42
team add validation to function processKeywordDetector
forge do:\n1. Add input validation\n2. Write tests\n3. Update README
Gated and redirected to blueprint:
forge fix this
autopilot build the app
team improve performance
forge add player trading
ultrawork make it better
Bypass the gate:
force: forge refactor the combat RemoteEvent handlers
! autopilot optimize everything
When the Gate Does Not Trigger
The gate auto-passes when it detects any concrete signal such as:
- file path
- issue or PR number
- function/class/symbol name
- explicit test runner target
- numbered steps
- acceptance criteria
- error reference
- provided code block
- escape prefix
End-to-End Flow Example
- User types
forge add ranked matchmaking.
- Gate detects an execution keyword plus an underspecified prompt.
- Gate redirects to blueprint with a message explaining the redirect.
- Blueprint consensus runs:
- Planner creates the initial plan
- Architect reviews the design
- Critic validates quality and testability
- On approval, the user chooses either:
$forge for sequential execution with verification
$team for parallel coordinated agents
- Execution begins with a clear, bounded plan.
Planning Checklist
- Identify the player-facing behavior and the creator outcome.
- Identify the Roblox runtime surfaces involved.
- Define who owns each mutation:
- server
- client
- shared module
- Define the canonical data shape using Luau-friendly structures.
- Define remote/event contracts only where they are truly needed.
- Define failure cases:
- stale replication
- duplicate grant
- invalid payload
- out-of-order UI updates
- partial save/load
- Define proof:
- which test
- which runtime path
- which state transition proves success
Anti-Slop Rules
- Do not emit architecture that reads like an enterprise CRUD app unless the repo actually contains that system shape.
- Do not use “DTO”, “controller”, “repository”, “microservice”, “middleware chain”, or similar language by default for Roblox gameplay code.
- Do not hand off to implementation until naming, ownership, and verification are concrete.
- Do not implement directly from blueprint. Approval hands off to
$forge or $team.
Follow-On Design Lanes
After the core blueprint, expand creator design depth when useful:
$blueprint:psych
$blueprint:loop
$blueprint:retention
$blueprint:social
Scenario Examples
Good: The user says continue after the workflow already has a clear next step. Continue the current planning branch instead of restarting discovery or re-asking the same question.
Good: The user changes only the downstream delivery shape, such as make a PR after approval. Preserve earlier non-conflicting planning constraints and update only the current handoff branch.
Bad: The user says continue, and the workflow restarts broad discovery or drifts into implementation before the plan and verification path are grounded.
Example
$blueprint "Plan a cross-server party queue for a Roblox combat experience.
Need:
- TeleportService handoff boundaries
- party ownership and invite data shape
- client/server split for queue UI
- anti-duplication rules for queue join/leave
- verification path for queue sync and teleport readiness"