AgentSkillsCN

backend-component-generator

以 content-ai 仓库的风格生成代码（Python、src 布局、click CLI、Pydantic 架构、确定性纯函数、pytest 覆盖率）。当您需要添加新模块、CLI 命令、流水线阶段，或队列功能时，可灵活运用此技能。

SKILL.md
--- frontmatter
name: backend-component-generator
description: Generates code in the content-ai repo style (Python, src layout, click CLI, Pydantic schemas, deterministic pure functions, pytest coverage). Use when adding new modules, CLI commands, pipeline stages, or queue features.

Backend Component Generator (content-ai style)

Name

Backend Component Generator

Description

You generate Python code matching this repo’s conventions: src/ layout, CLI via click, schema validation via Pydantic, deterministic pipeline stages, and tests via pytest.

Triggers

Use when the user asks:

  • •“Add a new module / feature”
  • •“Add a CLI command/flag”
  • •“Extend the queue system”
  • •“Implement new pipeline behavior”
  • •“Make it consistent with the repo style”

Instructions

Goal

Deterministic, typed, test-backed code with clean boundaries.

Conventions

  • •Code: src/content_ai/
  • •Tests: tests/
  • •Defaults: config/default.yaml
  • •cli.py parses/dispatches only; logic lives elsewhere.
  • •Pydantic for config + boundary payloads.
  • •IO/subprocess/network: clear errors, timeouts, no silent failures.
  • •Queue backend changes require tests + care (ACID-critical).

Output expectations

Include touched files + rationale + minimal test plan.

Constraints

No drive-by refactors. No new deps unless justified. Keep diffs tight.