| name | transversal-project-context-generator |
| description | Generates CONTEXT.md files for one or more company scenarios for a 4Geeks Academy transversal project milestone. A CONTEXT file is the company-specific document a student reads before implementing their milestone — it provides the real-world domain data (event schemas, KPIs, entity tables, seed instructions, restrictions) that turns a generic README into a concrete assignment. Requires an existing README for the milestone as input. Use this skill when adding a new company scenario to an existing milestone, when onboarding a new sector into the transversal project, or when generating context files after a README has already been written. Trigger on phrases like "generate the context for [company]", "add a new company to milestone N", "create CONTEXT files for this README", "generate CONTEXT for [sector] scenario", or when a user provides a README and one or more company descriptions. Do NOT use this skill to generate READMEs — use transversal-project-readme-generator for that. |
4Geeks Academy Transversal Project Context Generator
This skill generates company-specific CONTEXT.md files for transversal project milestones in the 4Geeks Academy AI Engineering bootcamp.
Repository Structure
Base Company References
The base company information and reference materials are located at:
https://github.com/4GeeksAcademy/course-outline-generator/tree/main/ai-engineering/project-milestone-context-base
This repository contains:
- Foundational company profiles
- Sector descriptions
- Domain data templates
- Reference schemas and examples
Always consult this repository when generating context files to ensure consistency and accuracy.
Output Repository
Generated CONTEXT files are saved to:
https://github.com/4GeeksAcademy/ai-engineering-sylllabus
Directory Structure:
- Folder format:
<two-digit-milestone>-<milestone-context>/
<two-digit-milestone>: Zero-padded milestone number (01, 02, 03, etc.)<milestone-context>: Descriptive context like "web-fundamentals", "automations", "training-rag", "backend-development"
- File format:
CONTEXT-<company>.md
<company>: Company/scenario name (retail, ecommerce, fintech, etc.)
Examples:
01-web-fundamentals/CONTEXT-retail.md
01-web-fundamentals/CONTEXT-ecommerce.md
02-automations/CONTEXT-fintech.md
03-training-rag/CONTEXT-healthcare.md
04-backend-development/CONTEXT-logistics.md
Overview
A CONTEXT file bridges the gap between the universal project statement (README) and the student's implementation. It provides:
- Real company scenario details
- Domain-specific event schemas and data structures
- Concrete KPIs and metrics relevant to the company
- Entity tables and seed data instructions
- Business constraints and requirements specific to that company
When to Use This Skill
Use this skill when:
- Adding a new company scenario to an existing milestone
- Onboarding a new sector/industry into the transversal project
- Generating context files after a README has been written
- A user provides a README and one or more company descriptions
- Updating or refreshing existing context files for a milestone
Do NOT use this skill to:
- Generate README files (use
transversal-project-readme-generator instead)
- Create the universal project statement
Required Inputs
-
Existing README.md for the milestone
- Must contain the universal project statement
- Should be sector-agnostic and generic
-
Milestone Context Information
- Milestone number (will be zero-padded to two digits)
- Milestone context name (e.g., "web-fundamentals", "automations")
-
Company Description(s)
- Can be one or more companies
- Should include sector, business model, and relevant domain details
- Reference the base company materials from the repository when available
Workflow
-
Receive Input
- Confirm you have the milestone README
- Confirm milestone number and context name
- Confirm you have company description(s)
- Fetch base company materials from repository if needed
-
Analyze the README
- Identify what needs to be made concrete
- Note any placeholders or generic references
- Understand the technical requirements
-
Research the Company Domain
- Consult base company references:
https://github.com/4GeeksAcademy/course-outline-generator/tree/main/ai-engineering/project-milestone-context-base
- Understand the sector-specific terminology
- Identify relevant data structures and events
-
Generate CONTEXT.md
- Create company-specific versions of generic concepts
- Provide concrete examples in the company's domain
- Include realistic seed data requirements
- Add domain-specific constraints
-
Structure for Repository
- Format milestone number as two digits (01, 02, etc.)
- Construct folder name:
<two-digit-milestone>-<milestone-context>/
- Name file as:
CONTEXT-<company>.md
- Full path:
<two-digit-milestone>-<milestone-context>/CONTEXT-<company>.md
-
Present to User
- Show the generated CONTEXT file
- Display the target repository path
- Explain how it connects to the README
- Confirm it aligns with the corresponding README
Output Format
Each CONTEXT.md file should include:
1. Company Introduction
- Company name and sector
- Brief business description
- Why this milestone matters to this company
2. Domain-Specific Data Structures
- Event schemas relevant to the company
- Data models and entity definitions
- Example data in the company's domain
3. Business Metrics and KPIs
- Specific metrics this company cares about
- How to calculate them
- Success thresholds
4. Seed Data Instructions
- What data the student should create
- Format and structure requirements
- Minimum data requirements
5. Business Constraints
- Company-specific rules and restrictions
- Edge cases to handle
- Domain-specific validations
6. Expected Deliverables
- What the student should produce
- How it should work for this specific company
- Acceptance criteria
Quality Standards
A good CONTEXT file should:
- ✅ Be completely self-contained (student shouldn't need to reference other docs)
- ✅ Use terminology authentic to the company's sector
- ✅ Provide concrete, realistic data examples
- ✅ Make all abstract concepts from the README concrete
- ✅ Be at the right difficulty level for the milestone
- ✅ Give students everything they need to implement without ambiguity
- ✅ Align with base company references when applicable
Example Interaction
User: "Generate the context for an e-commerce company for milestone 2 (automations)"
[Skill confirms:]
- Milestone number: 02
- Milestone context: "automations"
- Company: ecommerce
- Target path: 02-automations/CONTEXT-ecommerce.md
[Skill consults base company references]
[Skill generates CONTEXT-ecommerce.md with:]
- E-commerce events (purchase, cart_abandon, product_view)
- Automation triggers (low inventory, abandoned cart, price change)
- Metrics (conversion rate, automation success rate, time saved)
- Seed data for products, customers, automation rules
- E-commerce specific constraints (inventory thresholds, pricing rules)
Notes
- Always maintain consistency with the universal README
- The CONTEXT should NOT contradict or override the README's requirements
- Each company's CONTEXT can have different complexity levels if appropriate
- Students will use ONE CONTEXT file for their implementation
- Multiple CONTEXT files for the same milestone = multiple possible implementations
- Milestone context names should be kebab-case (lowercase with hyphens)
Repository Workflow
When generating files:
- Create the CONTEXT.md file with appropriate content
- Determine the folder structure:
<XX>-<milestone-context>/
- Name the file:
CONTEXT-<company>.md
- Inform the user of the complete path in the repository
- Optionally, if the user has repository access, offer to help prepare the file for commit
References
