Forwardimpact Monorepo Fit Pathway

Pathway Package

Web application, CLI, and formatters for career progression, job definitions, and agent profile generation. Two audiences use fit-pathway differently:

When to Use This Skill

CLI exploration and discovery:

• Exploring disciplines, tracks, levels, skills, behaviours, or drivers
• Generating or comparing job definitions across tracks and levels
• Understanding the proficiency and autonomy expected at each level
• Analyzing career progression gaps between current and target levels
• Generating AI agent configurations from the framework
• Answering questions about roles, skill expectations, or career paths

Framework setup and publishing:

• Setting up an organization's career framework project
• Building a static site for the framework
• Running a local development server to preview changes

How It Works

Job Derivation

A job is derived in real-time from three inputs: discipline, level, and optionally track. For each skill in the discipline:

1. The skill's tier determines its type — core (primary), supporting (secondary), or broad
2. The level's base proficiency for that type sets the starting point (e.g. "foundational" for primary skills at J060)
3. Track modifiers shift proficiency up or down per capability — a platform track with scale: +1 raises all skills in the scale capability by one level, while delivery: -1 lowers delivery skills
4. Positive modifiers are capped at the level's maximum base proficiency; results are clamped to the valid range (awareness → expert)

Behaviours follow the same pattern: base maturity from the level, then discipline and track modifiers stacked and clamped.

Skills from capabilities the track modifies positively — but that aren't in the base discipline — are added as broad-type "track-added" skills.

Agent Derivation

Agent profiles reuse job derivation with three additions:

1. Reference level is auto-selected — the first level where primary skills reach "practitioner", falling back to "working", then the middle level
2. Skill filtering removes isHumanOnly skills (physical presence, emotional judgment)
3. Skill focusing limits the matrix to the most relevant skills per stage, sorted by type (primary → secondary → broad → track-added)

Working styles are derived from the top behaviours by maturity — behaviours with positive track modifiers become personality traits that shape the agent's approach.

Tool Derivation

Tools come from toolReferences in skill definitions. Derivation collects all tool references from the job's highest-proficiency skills, deduplicates by name, and sorts alphabetically. Each tool tracks which skills reference it.

Interview Questions

Questions are selected from question banks organized by skill/behaviour ID, role type, and proficiency/maturity. Three interview types exist: mission fit (skill questions), decomposition (capability questions), and stakeholder simulation (behaviour questions). Selection prioritizes by skill type and fills a time budget.

Career Progression

Progression compares two job derivations (current vs target level) and calculates deltas — skill proficiency changes, gained/lost skills, and behaviour maturity changes.

Discovery Workflows

Common patterns for exploring the framework using the CLI. The CLI is data-driven — entity IDs depend on the YAML files in the active data directory. Start with summary commands to discover what's available.

Discover what roles exist on a track

# 1. See all tracks and which disciplines use them
npx fit-pathway track

# 2. See jobs filtered to a specific track
npx fit-pathway job --track=forward_deployed

# 3. List all job combinations on that track (for piping)
npx fit-pathway job --list --track=forward_deployed

# 4. View a specific role
npx fit-pathway job software_engineering J060 --track=forward_deployed

Understand a discipline's structure

# 1. See all disciplines with their type (professional/management) and valid tracks
npx fit-pathway discipline

# 2. Drill into a discipline to see skill tiers and behaviour modifiers
npx fit-pathway discipline software_engineering

# 3. Compare the same discipline across tracks
npx fit-pathway job software_engineering J060 --track=platform
npx fit-pathway job software_engineering J060 --track=forward_deployed

Explore career progression

# 1. See what changes between levels
npx fit-pathway progress software_engineering J060 --track=forward_deployed

# 2. Compare specific levels
npx fit-pathway progress software_engineering J060 --track=forward_deployed --compare=J100

Discover what a track modifies

# 1. View track detail to see all skill and behaviour modifiers
npx fit-pathway track forward_deployed

# 2. Compare the resulting skill matrices side by side
npx fit-pathway job software_engineering J060 --skills
npx fit-pathway job software_engineering J060 --track=forward_deployed --skills

CLI Reference

Getting Started

npx fit-pathway dev                           # Run live development server (default port 3000)
npx fit-pathway dev --port=8080               # Dev server on custom port
npx fit-pathway build --output=./public --url=https://example.com  # Static site
npx fit-pathway update                        # Update local ~/.fit/data/pathway/ installation
npx fit-pathway update --url=URL              # Update from custom source URL

init writes a starter framework into ./data/pathway/ (framework.yaml, levels.yaml, stages.yaml, drivers.yaml, plus disciplines/, capabilities/, behaviours/, and tracks/ directories). After init, validate with npx fit-map validate and explore with any of the entity commands below.

Entity Browsing

All entity commands support three modes:

Entity commands: discipline, level, track, behaviour, driver, stage, skill, tool.

Job Generation

npx fit-pathway job                                       # Summary with stats
npx fit-pathway job --track=<track>                       # Summary filtered by track
npx fit-pathway job --list                                # Valid combinations
npx fit-pathway job --list --track=<track>                # Combinations for a track
npx fit-pathway job <discipline> <level>                  # Trackless job
npx fit-pathway job <discipline> <level> --track=<track>  # With track
npx fit-pathway job <discipline> <level> --skills         # Skill IDs only
npx fit-pathway job <discipline> <level> --tools          # Tool names only

Agent Generation

npx fit-pathway agent --list                                        # Valid combinations
npx fit-pathway agent <discipline> --track=<track>                  # Preview
npx fit-pathway agent <discipline> --track=<track> --output=./agents # Write files
npx fit-pathway agent <discipline> --track=<track> --skills         # Skill IDs only
npx fit-pathway agent <discipline> --track=<track> --tools          # Tool names only

Interview, Progress & Questions

npx fit-pathway interview <discipline> <level>
npx fit-pathway interview <d> <l> --track=<t> --type=mission
npx fit-pathway progress <discipline> <level>
npx fit-pathway progress <d> <l> --compare=<to_level>
npx fit-pathway questions
npx fit-pathway questions --level=practitioner
npx fit-pathway questions --maturity=practicing
npx fit-pathway questions --skill=<id>
npx fit-pathway questions --behaviour=<id>
npx fit-pathway questions --capability=<id>
npx fit-pathway questions --stats
npx fit-pathway questions --format=yaml          # table, yaml, json

Interview types: mission (skill questions), decomposition (capability questions), stakeholder (behaviour questions).

Skill Detail

npx fit-pathway skill <id>          # Human-readable detail
npx fit-pathway skill <id> --agent  # Output as agent SKILL.md format

Agent Output Paths

• Agent profiles: .github/agents/{id}.agent.md (VS Code Custom Agents)
• Skill files: .claude/skills/{skill-name}/SKILL.md (Agent Skills Standard)

Data Resolution

The CLI resolves the data directory in this order:

1. --data=<path> flag (explicit override)
2. Upward traversal from CWD — looks for data/ (up to 3 parents)
3. ~/.fit/data/ (user-global fallback)

Use npx fit-map init to create a local ./data/ directory with starter framework data to get started.

Verification

npx fit-map validate    # Validate data files after changes
npx fit-pathway dev     # Preview changes in browser

Documentation

Before editing YAML framework data, read the relevant guide — they contain detailed examples, field references, and best-practice patterns essential for high-quality output.

• Authoring Frameworks Guide — How to write the YAML data: disciplines, levels, tracks, capabilities, skills, behaviours, stages, drivers. Includes proficiency vocabulary standards, co-located human/agent content patterns, checklist quality rules, and validation workflows
• Agent Teams Guide — How to generate, structure, and maintain exported agent teams. Covers the three-layer architecture (CLAUDE.md → agent profiles → skills), information flow rules, anti-patterns to avoid, and the maintenance checklist for reviewing exported output
• Career Paths Guide — Browse jobs, skills, and career progression between levels
• CLI Reference — Complete command reference for all Forward Impact CLI tools

Converted and distributed by TomeVault — claim your Tome and manage your conversions.