# Openspec Reused

> Create or update the reused.md artifact for an OpenSpec change — audit existing code for functionality to reuse, extend, or refactor before writing tasks. Use when planning implementation, after data-model and before tasks, or when the user mentions code reuse, DRY, duplicate logic, or extending existing components in an OpenSpec change.

- Skill: `vari-arti/openspec-reused` (Agent Skill)
- Install (CLI): `npx skillmds@latest add vari-arti/openspec-reused`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vari-arti/openspec-reused/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- License: MIT
- Author: vari-arti (https://skillmd.com/u/vari-arti)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/vari-arti/openspec-reused

---


Create the **reused.md** artifact for an OpenSpec change.

This document maps planned functionality to **existing code** so tasks extend or refactor instead of duplicating. Created **after data-model** and **before tasks**.

---

**Input**: Change name (kebab-case). If omitted, infer from conversation or list active changes.

**Steps**

1. **Select the change**
   - Use provided name, or infer from context
   - If ambiguous: `openspec list --json` and ask the user

2. **Verify prerequisites**
   ```bash
   openspec status --change "<name>" --json
   ```
   - `data-model`, `design`, and `specs` MUST be `done`
   - If not ready, stop and suggest completing dependencies first

3. **Read planning artifacts**
   - `proposal.md`, `design.md`, `data-model.md`, delta specs under `specs/`
   - Extract a checklist of concrete behaviors to implement (UI, API, BFF, enums, downloads, links, status mapping, etc.)

4. **Search the codebase** (for each checklist item)

   Use multiple strategies:
   - **SemanticSearch**: "Where is X implemented?", "How does Y work?"
   - **Grep**: class names, route paths, enum values, API endpoints from specs/design
   - **Similar patterns**: Link components (`*NameLink`), resource enrichers, `ImportClient` methods, status badges, download handlers, policy classes, form requests

   Typical locations in this project:
   - `app/Http/Controllers/`, `app/Services/`, `app/Clients/`
   - `resources/js/` (pages, components, hooks)
   - `routes/`, `app/Enums/`, `app/Policies/`
   - `tests/` for usage examples

5. **Performance reuse checklist (required)**

   For each feature or endpoint touched, explicitly check and record findings:
   - **N+1 queries**: resource/serializer calling repo per item, per-row DB lookups
   - **HTTP N+1**: per-item HTTP resolve (BFF or client) instead of batch/embed
   - **Polling chattiness**: frequent list/detail polling of heavy payloads
   - **Waterfall fetches**: sequential requests on a single page instead of parallel/aggregate
   - **Over-fetching**: list endpoints returning large fields unused by UI
   - **Duplicate work**: repeated computation/lookups across list → show flows

6. **Classify each finding**

   | Action | When |
   |--------|------|
   | **reuse** | Existing code satisfies the requirement with no change |
   | **extend** | Existing module/class is the right place to add behavior |
   | **refactor** | Two similar implementations exist — extract shared piece first |
   | **new** | No suitable code; justify in Gaps |

7. **Get artifact instructions**
   ```bash
   openspec instructions reused --change "<name>" --json
   ```
   Write to `resolvedOutputPath` using the template.

8. **Write reused.md**
   - One **Findings** block per spec requirement or design decision
   - Table rows: Action, Path, Symbol (`Class::method` or component name), Notes
   - **Recommendation** per block with clear next step
   - **Gaps** for items marked **new**
   - **Task Hints**: bullet list mapping reuse decisions → task wording (e.g. "Extend `SupplierNameLink` → create `PresetNameLink` with same props pattern")

9. **Validate**
   ```bash
   openspec status --change "<name>"
   openspec schema validate spec-driven
   ```

**Output**

Summarize:
- Change name and path to `reused.md`
- Reuse vs extend vs new counts
- Top 2–3 highest-impact reuse opportunities
- Next step: create `tasks.md` — tasks must reference paths from reused.md

**Guardrails**

- Every path must be real — verify files exist before listing
- Do not recommend **new** when **extend** is possible on an existing class
- Prefer extending BFF enrichers, shared components, and clients over parallel copies
- If design proposes new code but reuse audit finds equivalent — flag in Task Hints and suggest design update
- Write prose in Russian; keep paths, symbols, and enum literals in English as in code