---

name: fit-universe description: > Synthetic data generation CLI. Generates framework definitions, organizational documents, activity data, and personal knowledge base content from a universe DSL file. Use when generating example data, testing with synthetic datasets, or working with the universe DSL.

fit-universe CLI

Generate synthetic data for the entire Forward Impact suite from a single DSL file. The CLI orchestrates parsing, entity generation, optional LLM prose, and rendering into multiple output formats.

When to Use

• Generating example data for development or testing
• Creating synthetic pathway frameworks for new installations
• Producing organizational documents, activity records, and KB content
• Testing pipeline changes end-to-end
• Writing or editing universe DSL files

---

CLI Reference

npx fit-universe                     # Structural generation only (no LLM)
npx fit-universe --cached            # Use cached prose (fast, repeatable)
npx fit-universe --generate          # Generate prose via LLM (requires LLM_TOKEN)
npx fit-universe --cached --strict   # Fail on cache miss
npx fit-universe --load              # Load raw docs to Supabase Storage
npx fit-universe --only=pathway      # Render only one content type
npx fit-universe --dry-run           # Show what would be written
npx fit-universe --universe=path     # Custom universe file

Content Types

Use --only=<type> to generate a single content type:

| Type | Output Directory | Contents | | ---------- | ------------------------- | --------------------------------- | | html | examples/organizational | Articles, guides, FAQs, courses | | pathway | examples/pathway | YAML framework files | | raw | examples/activity | Roster, GitHub events, evidence | | markdown | examples/personal | Briefings, notes, KB content |

Prose Modes

| Mode | Flag | Description | | ---------- | -------------- | ---------------------------------------- | | no-prose | (default) | Structural only, no LLM calls | | cached | --cached | Read from .prose-cache.json | | generate | --generate | Call LLM, write to cache |

---

Universe DSL

Universe files define a complete synthetic environment. The default file is at libraries/libuniverse/data/default.dsl.

Top-Level Blocks

universe Name {
  domain "example.dev"
  industry "technology"
  seed 42

  org hq { ... }
  department engineering { ... }
  team backend { ... }
  people { ... }
  project alpha { ... }
  snapshots { ... }
  scenario launch_push { ... }
  framework { ... }
  content guide_html { ... }
  content basecamp_markdown { ... }
}

Key Blocks

org / department / team — Organizational hierarchy with headcounts, managers, and repo assignments.

people — Count, name theme, level distribution, discipline distribution.

project — Cross-team initiatives with timelines and prose topics.

snapshots — GetDX snapshot generation (quarterly intervals).

scenario — Time-bounded effects on teams (commit volume, DX driver trajectories, evidence generation).

framework — Full pathway framework: levels, capabilities with skills, behaviours, disciplines with skill tiers, tracks, drivers, and stages.

content — Output content blocks specifying article/blog/FAQ counts, persona configurations, and briefing counts.

---

Data Resolution

The production universe DSL lives at libraries/libuniverse/data/universe.dsl. The default test universe is libraries/libuniverse/data/default.dsl. Use --universe=path to specify a custom file.

All generated output writes to examples/ at the monorepo root.

---

Environment

Generation requires LLM_TOKEN and LLM_BASE_URL when using --generate mode. These are always available in the standard environment (see AGENTS.md).

npx fit-universe --generate          # Uses LLM_TOKEN from environment

Verification

After generation, the CLI runs cross-content validation automatically and reports pass/fail for each check. Validate the generated pathway data separately:

npx fit-map validate --data=examples/pathway