project-type-determination

Project Type Determination

Determine the project type to configure appropriate analysis thresholds for the python-plan-optimization skill.

Core Mission

Identify the project type through a 3-step workflow:

1. Check if type was explicitly provided in the invoking prompt
2. Scan plan files for explicit project type statements
3. Ask the user if type cannot be determined

Deliverable: Project type classification with source attribution.

Supported Project Types

Type	Key Characteristic	Analysis Mode
poc	Proof of concept, throw-away code	Critical issues only
mvp	Minimum viable product, ship fast	Critical + High issues
private	Personal/learning project	All issues, educational
enterprise	Production team project	All issues, verification required
opensource	Public library/tool	All issues, types + docs required

Workflow

Step 1: Check Invoking Prompt

Look for explicit project type in the user's request:

Direct Mentions:

• "Analyze my enterprise project's PLAN.md"
• "Review this MVP plan"
• "Check my POC code in design.md"
• "Review my open source library"
• "Analyze my personal project"

Pattern Matching:

Pattern	Detected Type
enterprise, production	enterprise
mvp, minimum viable	mvp
poc, proof of concept, prototype, spike	poc
open source, public library, pypi, pip install	opensource
personal, private, learning, experiment	private

If type found in prompt:

project_type: [detected type]
source: prompt

→ Done. Skip to Output.

If not found: Continue to Step 2.

---

Step 2: Scan Documents for Explicit Statements

Read the target documents and search for explicit project type declarations.

Explicit Statement Patterns:

Type	Patterns to Match
poc	"This is a (POC|proof of concept|prototype)", "POC for...", "Spike to validate..."
mvp	"MVP (for|development|plan)", "Minimum viable product", "MVP plan for..."
private	"Personal project", "Private project", "Learning exercise", "Experiment to..."
enterprise	"Enterprise (system|deployment|project)", "Production (system|deployment)", "Enterprise-grade"
opensource	"Open source (library|package|project)", "Public (library|package|API)", "PyPI package", "Published to PyPI", "pip install"

Search Strategy:

1. Use Grep to search document(s) for type patterns
2. Look in document headers, introductions, and project descriptions
3. Check for explicit statements like "This is a plan for an MVP"

If explicit statement found:

project_type: [detected type]
source: document

→ Done. Skip to Output.

If not found: Continue to Step 3.

---

Step 3: Ask User

When project type cannot be determined from context, ask the user.

First, output explanation text:

I could not determine the project type from the prompt or documents. The project type affects analysis depth:

• POC: Critical issues only (throw-away code)
• MVP: Critical + High (ship fast, track debt)
• Private: All suggestions (learning focus)
• Enterprise: Strict analysis, verification required
• Open Source: API stability, required types & docs

Then invoke AskUserQuestion:

questions:
  - question: "What type of project is this?"
    header: "Project Type"
    multiSelect: false
    options:
      - label: "Private (Recommended)"
        description: "Full analysis with all suggestions - best for learning and personal projects"
      - label: "Open Source"
        description: "API-focused with required type hints and documentation"
      - label: "Enterprise"
        description: "Strict analysis with verification required for all claims"
      - label: "MVP"
        description: "Balanced analysis - critical and high priority issues only"
      - label: "POC"
        description: "Minimal analysis - only critical blocking issues"

Map user selection to type:

Selection	Type
Private (Recommended)	private
Open Source	opensource
Enterprise	enterprise
MVP	mvp
POC	poc
Other (custom input)	Parse or default to private

After user responds:

project_type: [selected type]
source: user

---

Output Format

Always output in this format for the optimization skill to parse:

project_type: [poc|mvp|private|enterprise|opensource]
source: [prompt|document|user]

Example Outputs:

# Type found in prompt
project_type: enterprise
source: prompt

# Type found in document
project_type: mvp
source: document

# User selected type
project_type: private
source: user

Constraints

DO:

• Check prompt first (fastest path)
• Use exact pattern matching for document scanning
• Present explanation before asking user
• Default to "private" for ambiguous "Other" responses

DO NOT:

• Guess project type from code characteristics
• Infer type from file structure or naming
• Make assumptions about team size or deployment
• Skip the user question when type is ambiguous

Success Criteria

Skill succeeds when:

• Project type is determined through one of three methods
• Output contains valid project_type and source
• User interaction (if needed) provides clear explanation before question