Design and document robust APIs with clear endpoints, fostering developer collaboration and scalability.

2.0

Anthropic: Claude 3.5 Sonnet

Jira issue (full context) Project: Purpose Project: Features / Advantages Project: Target Audience Project: Tech Stack

• Input: A document describing the game mechanics from contexts

• AI Task

  • Extract key aspects: rules, user flows, data flows, core functionality, etc.

  • Create a glossary of relevant terms and definitions.

  • Determine roles/entities involved (e.g., user, admin, NPC, etc.).

• Output: A concise, structured overview covering:

  • Fundamental concepts and components

  • Main interactions or workflows

  • Basic relationships among these elements

• Input: Output from "Requirements Analysis" (glossary, core components, etc.)

• AI Task:

  • Compile a full list of possible user actions (in a game context: movement, item usage, etc.; in software: CRUD actions, triggers, etc.).

  • For each action:

    • Name (short tag)

    • Short description

    • Requirements/conditions (e.g., parameters, permissions, items)

    • Result (changes in system state, UI updates, etc.)

• Output: Comprehensive list (table or hierarchy) of all user actions with detailed descriptions

• Input: User actions and entities

• AI Task:

  • Identify main entities (e.g., Player, User, Item, Resource, etc.) and their fields (id, status, relationships, etc.).

  • Determine client-side vs. server-side storage (or local vs. cloud for software).

  • Define how and when the system synchronizes data.

• Output: A schema (ER diagram or tabular format) clarifying how data is structured and updated

• Input: User actions and the data model

• AI Task:

  • Map endpoints or service methods to each action (REST, GraphQL, WebSocket, etc.).

  • Specify request/response formats (HTTP method, path, parameters, body, etc.).

• Output: A table with next columns:

  • Endpoint: API route.

  • HTTP Method: GET, POST, PUT, DELETE.

  • Request Body: JSON schema for the request payload (use `json` formatting).

  • Response Structure: JSON schema for the response payload (use `json` formatting).

• Input: Draft API specification and the list of user actions

• AI Task:

  • Verify that all actions are covered (no duplicates or gaps).

  • Adjust data structures or endpoints if needed.

• Output: A refined API spec addressing all requirements and potential concerns