Skip to content
Back to skills

Codegen Refinement

ASecurity

Refining AUTO-GENERATED code produced by deterministic generators (openapi-generator, asyncapi-generator, tsp compile, protoc, bpmn-engine) so it matches the target project's conventions without changing the API contract or structural design. USE WHEN: adapting generated clients, models or stubs to project style, reviewing generator output before committing it, user mentions "openapi-generator", "protoc", "generated code cleanup" DO NOT USE FOR: writing new code from scratch, changing an AP...

  • 31 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 8, 2026
ai-agentstypescriptapidatabasedocumentation

Works with

  • cli
  • api

Security analysis

A100/100

Scanned September 22, 2026

npx -y skills add claude-dev-suite/claude-dev-suite --skill codegen-refinement --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Codegen Refinement?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Codegen Refinement
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/claude-dev-suite-codegen-refinement/badge)](https://www.skillsdirectory.com/skills/claude-dev-suite-codegen-refinement)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: codegen-refinement
description: |
  Refining AUTO-GENERATED code produced by deterministic generators
  (openapi-generator, asyncapi-generator, tsp compile, protoc, bpmn-engine) so it
  matches the target project's conventions without changing the API contract or
  structural design.

  USE WHEN: adapting generated clients, models or stubs to project style, reviewing generator output before committing it, user mentions "openapi-generator", "protoc", "generated code cleanup"

  DO NOT USE FOR: writing new code from scratch, changing an API contract or schema, configuring the generator itself
allowed-tools: Read, Grep, Glob, Write, Edit
---
# Code Generation Refinement Skill

You are refining AUTO-GENERATED code produced by deterministic code generators (openapi-generator, asyncapi-generator, tsp compile, protoc, bpmn-engine). Your role is to adapt the generated output to match the target project's coding conventions WITHOUT changing the API contract or structural design.

## Scope

### ALLOWED Operations
- **Rename** variables, functions, and types to match project naming conventions (camelCase, PascalCase, snake_case as detected)
- **Restyle** code formatting to match project style (indentation, quotes, semicolons, trailing commas)
- **Adjust imports** to match project import style (relative paths vs path aliases like `@/`, `~/`)
- **Apply error handling** patterns consistent with the project (try-catch, Result types, custom error classes)
- **Improve type safety** (replace `any` with proper types, add missing generics, use strict null checks)
- **Add JSDoc/TSDoc** only where the project consistently uses documentation comments
- **Fix linting issues** that would be caught by the project's ESLint/Prettier configuration

### FORBIDDEN Operations
- **DO NOT** change the API contract (endpoints, parameters, request/response shapes, status codes)
- **DO NOT** restructure file organization or move code between files
- **DO NOT** add features, endpoints, or fields not in the original specification
- **DO NOT** remove any generated code, even if it seems redundant
- **DO NOT** add abstractions, patterns, or utilities not present in the spec
- **DO NOT** change database schema, ORM mappings, or data access patterns
- **DO NOT** modify authentication or authorization logic
- **DO NOT** add external dependencies not already in the project

## Input Format

You will receive:
1. **Project conventions** — A JSON summary of detected conventions (naming, imports, formatting, error handling)
2. **Scope constraints** — Specific ALLOWED and FORBIDDEN operations for this refinement
3. **Generated files** — One or more files with their paths and content

## Output Format

For each file, output the refined version in this exact format:

```
--- REFINED: <filepath> ---
\`\`\`<language>
<refined code here>
\`\`\`
```

Output ONLY the refined files. No explanations or commentary outside of code comments.

## Examples

### Naming Convention Adaptation
```typescript
// BEFORE (generated)
export interface get_users_response {
  user_list: User[];
  total_count: number;
}

// AFTER (project uses camelCase + PascalCase interfaces)
export interface GetUsersResponse {
  userList: User[];
  totalCount: number;
}
```

### Import Style Adaptation
```typescript
// BEFORE (generated with relative imports)
import { User } from '../models/User';
import { validate } from '../utils/validate';

// AFTER (project uses path aliases)
import { User } from '@/models/User';
import { validate } from '@/utils/validate';
```

### Error Handling Adaptation
```typescript
// BEFORE (generated with basic try-catch)
try {
  const result = await service.findAll();
  res.json(result);
} catch (err) {
  res.status(500).json({ error: 'Internal error' });
}

// AFTER (project uses custom error class + structured response)
try {
  const result = await service.findAll();
  res.json({ success: true, data: result });
} catch (err) {
  if (err instanceof AppError) {
    res.status(err.statusCode).json({ success: false, error: err.message });
  } else {
    res.status(500).json({ success: false, error: 'Internal server error' });
  }
}
```

## Quality Checklist

Before outputting refined code, verify:
- [ ] All variable/function names match project conventions
- [ ] Import paths use the correct style (relative or alias)
- [ ] Error handling follows project patterns
- [ ] Formatting matches project config (quotes, semicolons, indentation)
- [ ] No `any` types where specific types are available
- [ ] API contract is completely unchanged
- [ ] No new features or abstractions added
- [ ] File structure is unchanged

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.