recipe-implement

name	recipe-implement
description	Orchestrate the complete implementation lifecycle from requirements through verified repository implementation.

Required Skills [LOAD BEFORE EXECUTION]

1. [LOAD IF NOT ACTIVE] subagents-orchestration-guide — agent coordination and workflow flows
2. [LOAD IF NOT ACTIVE] documentation-criteria — scale-selected document path
3. [LOAD IF NOT ACTIVE] requirement-convergence — outcome, exclusion, and rough-cost convergence before design
4. [LOAD IF NOT ACTIVE] llm-friendly-context — cross-agent handoffs and task carrier

Spawn rule: every spawn_agent call uses fork_turns="none" so the subagent receives only the task message and explicitly provided context.

Full-Cycle Implementation

$ARGUMENTS

Orchestrator Definition

Core Identity: Coordinate the lifecycle, complete lightweight workflow operations directly, and invoke named specialists for their domain work.

Follow the scale-selected flow and its user approval points from subagents-orchestration-guide.

Step 1: Requirement Analysis

Spawn requirement-analyzer for compact request signals, scope evidence, cost evidence, affected-layer evidence, and decision-changing questions.

At the requirements stop, the orchestrator applies subagents-orchestration-guide Requirement Convergence from the user's wording and supplied evidence, resolves material questions, determines Structural Scale and affected layers, and selects the canonical route.

[STOP — BLOCKING] Present the converged requirement record, scale, affectedLayers, and scope to the user for confirmation. CANNOT proceed until user explicitly confirms.

Step 2: Canonical Workflow Routing

Apply the subagents-orchestration-guide Basic Flow for Work Planning using scale as the primary route. Use affectedLayers only to select the layer-specific additions and agents within that route:

affectedLayers	Layer-specific routing
["backend"] only	Backend agents
["frontend"] only	UI Spec when required by the canonical flow, then frontend designer, executor, and quality fixer
["backend", "frontend"]	Fullstack monorepo flow with layer-specific analysis, design, and task routing

The scale-selected Large, Medium, or Small flow remains authoritative for document creation, codebase analysis, design verification, planning, and execution.

Autonomous Execution Mode

Enter autonomous execution when the subagents-orchestration-guide Authority Grant is satisfied.

Per-Task Execution Cycle

For a fullstack task set, apply the Fullstack Flow filename routing exposed by subagents-orchestration-guide. For a single-layer task set, use the executor and quality fixer selected by affectedLayers. Execute each task through the canonical autonomous task cycle.

Post-Implementation Review (After All Tasks Complete)

Apply subagents-orchestration-guide Post-Implementation Review to the actual repository changes and governing documents. Use the active Small task file as the governing source when no durable document exists. Resolve required fixes through the normal task cycle.

Apply a security-reviewer finding only when leaving it unresolved would violate an explicit governing requirement or repository rule, or leave a concrete material security failure in the actual reachable trust model. The violated requirement, rule, or failure defines implementation scope: route the smallest correction that resolves it, treating the reviewer's suggestion as one candidate implementation.

Test Information Communication

Verify acceptance-test-generator artifact paths and pass them to work-planner.

Completion Criteria

• Scope/cost evidence was collected and the orchestrator's requirement and scale decisions were user-confirmed
• Layer routing determined (backend / frontend / fullstack)
• Correct workflow followed per layer routing
• codebase-analyzer included before Design Doc creation for Medium/Large flows
• code-verifier discrepancies passed through Review Resolution before Design Doc review
• All stopping points honored with user confirmation obtained
• Quality-fixer spawned before every commit
• All tasks committed or user input requested