Package Exports

@forwardimpact/pathway
@forwardimpact/pathway/derivation
@forwardimpact/pathway/interview
@forwardimpact/pathway/levels
@forwardimpact/pathway/loader
@forwardimpact/pathway/matching
@forwardimpact/pathway/types
@forwardimpact/pathway/validation

Readme

Engineering Pathway

A unified framework for human and AI collaboration in engineering. Define roles, track skills and behaviours, build career paths, and generate AI coding agents—all from the same coherent foundation.

Quick Start

Install as a Package

# Create a new project
mkdir my-org-pathway && cd my-org-pathway
npm init -y
npm install @forwardimpact/pathway

# Initialize with example data
npx pathway init

# Validate data
npx pathway --validate

# Start web app
npx pathway serve

Open http://localhost:3000 to explore jobs, skills, and behaviours.

For Development

Clone the repository and run locally:

git clone https://github.com/forwardimpact/pathway.git
cd pathway
npm install
npm start

Customization

Edit files in ./data/ to match your organization:

grades.yaml — Career levels
disciplines/*.yaml — Engineering disciplines
tracks/*.yaml — Role tracks (platform, frontend, etc.)
skills/*.yaml — Technical skills with level descriptions
behaviours/*.yaml — Behavioural expectations

CLI Reference

# Getting started
npx pathway init                      # Create ./data/ with examples
npx pathway serve --port=8080         # Serve web app

# Browse data
npx pathway skill                     # Summary of all skills
npx pathway skill --list              # Skill IDs for piping
npx pathway job software_engineering platform senior

# Generate outputs
npx pathway interview software_engineering platform L4
npx pathway agent software_engineering platform --output=./.github/agents

# Validation
npx pathway --validate                # Validate all data files
npx pathway --help                    # Full command reference

API Usage

import { loadAllData } from "@forwardimpact/pathway";
import { deriveJob } from "@forwardimpact/pathway/derivation";

const data = await loadAllData("./data");
const job = deriveJob({
  discipline: data.disciplines[0],
  track: data.tracks[0],
  grade: data.grades[0],
  skills: data.skills,
  behaviours: data.behaviours,
});

Documentation

Overview — Overview of all topics
Core Model — Disciplines, Tracks, Grades, Skills, Behaviours, Capabilities, Drivers, and Job Derivation
Lifecycle — Stages, handoffs, and checklists (applies to both humans and agents)
Agents — Agent profile derivation, SKILL.md format, and VS Code integration
Reference — File organization, templates, and CLI usage

Development

npm run check      # Format, lint, and test
npm run validate   # Validate YAML data files
npm run test:e2e   # Run end-to-end tests