Phase 1: Understand the Task

Phase 1:

Understand the Task

What Is This?

“Phase 1: Understand the Task” is the initial step in the estimation process for the Happycapy Skills platform’s estimate skill. This phase requires a careful and structured reading of the task description provided as an argument to the skill. Its primary objective is to ensure that the task is fully understood and clearly defined before any attempt is made to estimate the required effort. If the task description lacks sufficient detail or contains ambiguities, clarification should be requested from the task author. This prevents inaccurate estimates and reduces the risk of misunderstandings later in the development cycle.

Why Use It?

A reliable estimate depends on a thorough understanding of what needs to be done. Vague or incomplete task descriptions are a leading cause of estimation errors, missed requirements, and scope creep. By enforcing a disciplined review of the task description, this phase ensures that all parties share a common understanding of the work to be performed. This not only increases the accuracy of subsequent estimates but also promotes better communication and alignment within the team.

Moreover, the skill mandates reviewing key project documentation, such as the CLAUDE.md file and relevant design documents. This contextual information provides insight into the project’s technical landscape, coding standards, and architectural constraints, all of which are crucial for forming a realistic estimate.

How to Use It

When the estimate skill is invoked, it expects a task description as its argument:

estimate "Implement a new save/load system for player profiles"

Upon receiving the argument, the skill should:

1. Read and Analyze the Task Description:

   • Parse the provided description for clarity and specificity.
   • Identify the main deliverables, acceptance criteria, and any explicit constraints or dependencies.
   • Example of a clear task:
     "Add leaderboard integration to the main menu, showing top 10 players from the online backend."
   • Example of a vague task:
     "Improve main menu."

2. Request Clarification if Needed:

   • If the description is too vague or ambiguous to estimate meaningfully, the skill must immediately respond with a request for additional information.
   • Example response:
     "The task description 'Improve main menu' is too vague to estimate. Please specify which features, UI elements, or performance aspects should be improved."

3. Review Project Context:

   • Read CLAUDE.md to gather information about:
     • Tech stack (e.g., Python, Unity, React)
     • Coding standards (e.g., code style, documentation requirements)
     • Architectural patterns (e.g., MVC, event-driven)
     • Estimation guidelines (e.g., story points, t-shirt sizing)
   • If the task relates to a documented feature, read relevant design documents under design/gdd/ to understand requirements, edge cases, and integration points.

Example Workflow

Suppose the skill receives:

estimate "Add cloud save support"

If the CLAUDE.md indicates prior work on save systems and a design doc exists at design/gdd/save_system.md, the skill should:

• Check if “cloud save” is defined in the design doc.
• If the term is ambiguous (e.g., not specifying which cloud provider or user authentication method), ask:
  "Please clarify which cloud service should be used (e.g., AWS, Google Cloud) and whether user authentication is required."

When to Use It

• Every time a new task is submitted for estimation:
  Phase 1 should always be performed at the outset, regardless of how familiar the team may be with the project or the author of the task.
• When requirements change:
  If requirements evolve or new information surfaces, re-running Phase 1 ensures that the updated scope is well understood before re-estimating.
• During backlog grooming and sprint planning:
  Applying this phase during planning sessions helps surface unclear tasks before they block development progress.

Important Notes

• Do not proceed to estimation on vague tasks:
  If the task description lacks sufficient detail, halt the process and request clarification. Never attempt to guess or assume missing information.
• Project context is mandatory:
  Always read CLAUDE.md and relevant design documentation before attempting any estimate, as these resources may contain critical information affecting scope and complexity.
• Follow structured communication:
  When requesting clarification, be specific about what details are missing and, if possible, provide examples of the required information.
• Automated tool support:
  The skill is permitted to use tools like Read, Glob, and Grep to retrieve and parse documentation efficiently, but it should never “invent” requirements.

Code Example for Clarification Request

def request_clarification(task_description):
    return (
        f"The task description '{task_description}' is too vague to estimate. "
        "Please provide specific details, such as functionality, scope, or acceptance criteria."
    )

## Usage
desc = "Optimize performance"
if is_too_vague(desc):
    print(request_clarification(desc))

By rigorously applying “Phase 1: Understand the Task,” the Happycapy Skills platform ensures all subsequent estimation phases are built on a foundation of clarity and shared understanding, reducing risk and improving overall productivity.