- name
- recipe-fullstack-implement
- description
- Run the full-cycle implementation workflow for one outcome spanning backend and frontend layers.
## Required Skills [LOAD BEFORE EXECUTION]
Before loading required skills, use `wc -c` to size the files and set both `exec_command.max_output_tokens` and the outer `// @exec: {"max_output_tokens": N}` budget to accommodate their contents, including the combined output of batched reads. Split oversized reads upfront; recover any truncated content before proceeding.
1. [LOAD IF NOT ACTIVE] `subagents-orchestration-guide` — Fullstack (Monorepo) Flow, approvals, and autonomous execution
2. [LOAD IF NOT ACTIVE] `documentation-criteria` — scale-selected document path
3. [LOAD IF NOT ACTIVE] `requirement-convergence` — outcome, exclusions, and rough-cost challenge
4. [LOAD IF NOT ACTIVE] `llm-friendly-context` — cross-agent handoffs and Small task carrier
**Invocation rule**: Start each named agent with `spawn_agent` and `fork_turns="none"`. When invoking the same named agent again within this flow, use `followup_task` on its existing instance with the inputs specified for the current invocation. Supply only the exact artifacts needed by that specialist.
Requirements or continuation instruction: $ARGUMENTS
## Entry
- New or scope-changing requirements: apply the subagents-orchestration-guide Requirement Evidence Handoff and invoke requirement-analyzer for compact scope/cost evidence; user selections establish requirements and exclusions, while the orchestrator judges readiness and determines scale and layer routing.
- Existing PRD, UI Spec, Design Docs, Work Plan, or tasks: resume at the next incomplete Fullstack (Monorepo) Flow phase. Restart requirement analysis when the approved outcome, requirement, or exclusion changes.
- Quality failure during an existing implementation: resume its task cycle and Orchestrator Escalation Resolution.
For a requirements stop, read `requirement-convergence`'s `references/scope-confirmation.md` and render that output from the retained user record and analyzer evidence. Render each unanswered product, UX, or operational decision under **User decisions** as an unresolved question followed by its answer-dependent effects. Only an explicit user selection moves it into **Confirmed scope**. Record Structural Scale and workflow routing under **Workflow**.
Resolve the entry from supplied artifacts and repository state. Ask only when different interpretations require a change to user-required outcomes, explicit constraints, or execution authority.
## Flow
Apply `subagents-orchestration-guide`'s `references/monorepo-flow.md` with backend, frontend, and shared routing. The orchestrator directly owns artifact/path resolution, execution-plan updates, approval recording, task-set computation, commits, and lightweight checks; invoke the named specialists for analysis, authoring, implementation, review, and quality judgment.
Reuse the active execution plan or register the material remaining phases once. Follow the Fullstack (Monorepo) Flow's document approvals. After implementation-scope approval, execute tasks autonomously through its filename routing, Per-Task Change Set, quality gate, commit, and Post-Implementation Review.
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.
External evidence, prototypes, and repository environment preparation remain conditional under the canonical flow. A missing optional input does not create a stop.
## Completion Check
- The existing workflow state was resumed rather than duplicated.
- Backend, frontend, and shared work used the Fullstack (Monorepo) Flow routing.
- Required document approvals and quality-before-commit gates were preserved.
- After implementation approval, work stopped only for a genuine user-owned escalation condition.
- Final code/security verification and completion reporting finished.
View on GitHub