Constraint-based schedule solver -- define tasks, dependencies, and resources; get a timed plan.
$ npx skejj make examples/roast-chicken.json
Roast Chicken Dinner
A classic Sunday roast with oven-constrained cooking steps. The oven (capacity 1) forces chicken and potatoes to roast sequentially.
17:00 17:30 18:00 18:30 19:00
─ Mains ───────────────────────────────────────────────────────────────────────
Prep chicken ████████ │ │ │
Roast chicken │ █████████████████████████████████████████████
Make gravy │ │ │ │ █████
Plate up │ │ │ │ ███
─ Sides ───────────────────────────────────────────────────────────────────────
Prep potatoes ░░░░░░░░░░ │ │ │
Steam vegetables ░░░░░░░░ │ │ │
Roast potatoes │ │ │ ░░░░░░░░░░░░░░░░░░░░
─ Setup ───────────────────────────────────────────────────────────────────────
Set table ░░░░░ │ │ │
Preheat Oven │ ░░░ │ │ │
--- Summary ---
Total time: 2h
Steps: 9
Critical path: 2h (4 steps)
Resources used: Oven
--- Critical Path ---
Prep chicken -> Roast chicken -> Make gravy -> Plate up
Float per step:
Prep chicken: 17:00 -> 17:15 [0m (critical)]
Prep potatoes: 17:00 -> 17:20 [55m slack]
Steam vegetables: 17:00 -> 17:15 [100m slack]
Set table: 17:00 -> 17:10 [105m slack]
Preheat Oven: 17:10 -> 17:15 [10m slack]
Roast chicken: 17:15 -> 18:45 [0m (critical)]
Roast potatoes: 18:15 -> 18:55 [55m slack]
Make gravy: 18:45 -> 18:55 [0m (critical)]
Plate up: 18:55 -> 19:00 [0m (critical)]
npx skejj make examples/roast-chicken.jsonNo install needed -- npx downloads and runs skejj directly.
To install globally:
npm install -g skejjThen use skejj instead of npx skejj everywhere.
Start from scratch with the interactive wizard:
skejj newGenerate a schedule from a description using AI:
skejj generate "plan a birthday party for 20 kids"Solve a schedule file and display the timed plan as an ASCII Gantt chart.
Critical-path steps are shown with solid blocks (█). Non-critical steps with lighter blocks (░). The summary includes total time, step count, critical path length, and float per step.
| Flag | Default | Description |
|---|---|---|
-o, --output <file> |
stdout | Write output to file instead of stdout |
-q, --quiet |
false | Suppress summary stats, show only the schedule |
-f, --format <type> |
-- | Export format: gantt, csv, json (writes a file in addition to ASCII output) |
--width <cols> |
terminal width | Chart width in columns (default: terminal width or 80) |
# Solve and display in the terminal
skejj make examples/roast-chicken.json
# Export to CSV alongside ASCII output
skejj make myplan.json --format csv
# Write ASCII output to a file, suppress summary
skejj make myplan.json --quiet --output schedule.txt
# Export JSON data with a fixed chart width
skejj make myplan.json --format json --width 120Validate a schedule file without solving it. Reports schema errors and constraint warnings.
| Flag | Default | Description |
|---|---|---|
-q, --quiet |
false | Show only errors, suppress warnings |
# Validate and show all errors and warnings
skejj check myplan.json
# Show errors only (useful in CI)
skejj check myplan.json --quietGuided interactive wizard to create a new schedule from scratch. Prompts for schedule name, steps, durations, dependencies, and resources. Requires an interactive terminal (TTY).
skejj newGenerate a schedule JSON file from a natural language description using an LLM. Requires an API key configured via skejj config set.
| Flag | Default | Description |
|---|---|---|
-o, --output <file> |
derived from schedule name | Override output filename |
-f, --format <type> |
-- | Export format in addition to ASCII Gantt: gantt (Mermaid), csv, json
|
# Generate and display a birthday party schedule
skejj generate "plan a kids birthday party for 20 children"
# Generate and save to a specific file
skejj generate "weekend trip to Paris" --output paris-trip.json
# Generate and export a CSV
skejj generate "home renovation project" --format csvManage LLM provider configuration for the generate command.
Subcommands:
Set a configuration value. Supported keys: provider (e.g. openai, anthropic), apiKey, model.
Display current configuration.
# Set your LLM provider and API key
skejj config set provider openai
skejj config set apiKey sk-...
# Use Anthropic instead
skejj config set provider anthropic
skejj config set apiKey sk-ant-...
# Show current config
skejj config showInteractively adjust a solved schedule in a re-solve loop. Opens an interactive editor (requires TTY) where you can modify step durations, dependencies, and resources, and see the Gantt chart update live.
skejj adjust examples/roast-chicken.json
# Or adjust a schedule you created
skejj adjust myplan.jsonDemonstrates: equipment resource constraint (Oven, capacity 1 -- chicken and potatoes cannot roast simultaneously), backward scheduling from a target dinner time, and a linear dependency chain (prep -> roast -> gravy -> plate).
$ npx skejj make examples/roast-chicken.json
Roast Chicken Dinner
A classic Sunday roast with oven-constrained cooking steps. The oven (capacity 1) forces chicken and potatoes to roast sequentially.
17:00 17:30 18:00 18:30 19:00
─ Mains ───────────────────────────────────────────────────────────────────────
Prep chicken ████████ │ │ │
Roast chicken │ █████████████████████████████████████████████
Make gravy │ │ │ │ █████
Plate up │ │ │ │ ███
─ Sides ───────────────────────────────────────────────────────────────────────
Prep potatoes ░░░░░░░░░░ │ │ │
Steam vegetables ░░░░░░░░ │ │ │
Roast potatoes │ │ │ ░░░░░░░░░░░░░░░░░░░░
─ Setup ───────────────────────────────────────────────────────────────────────
Set table ░░░░░ │ │ │
Preheat Oven │ ░░░ │ │ │
--- Summary ---
Total time: 2h
Steps: 9
Critical path: 2h (4 steps)
Resources used: Oven
--- Critical Path ---
Prep chicken -> Roast chicken -> Make gravy -> Plate up
Float per step:
Prep chicken: 17:00 -> 17:15 [0m (critical)]
Prep potatoes: 17:00 -> 17:20 [55m slack]
Steam vegetables: 17:00 -> 17:15 [100m slack]
Set table: 17:00 -> 17:10 [105m slack]
Preheat Oven: 17:10 -> 17:15 [10m slack]
Roast chicken: 17:15 -> 18:45 [0m (critical)]
Roast potatoes: 18:15 -> 18:55 [55m slack]
Make gravy: 18:45 -> 18:55 [0m (critical)]
Plate up: 18:55 -> 19:00 [0m (critical)]
Demonstrates: people resource constraint (4 helpers shared across parallel tasks), forward scheduling from a party start time, and parallel tracks running simultaneously (food, venue setup, activities).
$ npx skejj make examples/birthday-party.json
Kids Birthday Party
Planning a kids birthday party with a team of helpers. People resource allocation ensures we never exceed 4 helpers simultaneously.
10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:3013:40
─ Food ────────────────────────────────────────────────────────────────────────
Bake cake ░░░░░░░░░░░░░░░░░│ │ │ │ │ │
Ice cake │ │ ░░░░░ │ │ │ │ │
Serve cake │ │ │ │ │ │ ████ │
─ Venue ───────────────────────────────────────────────────────────────────────
Buy supplies █████████████████│ │ │ │ │ │
Decorate venue │ │ ████████████ │ │ │ │
Blow up balloons │ │ │ │ ░░░░░░░░ │ │ │
Clean up │ │ │ │ │ │ │ ████████
─ Activities ──────────────────────────────────────────────────────────────────
Welcome guests │ │ │ │ ███ │ │ │ │
Party games │ │ │ │ █████████████████ │ │
Set up games │ │ │ │ │ ░░░░░░ │ │
--- Summary ---
Total time: 3h 40m
Steps: 10
Critical path: 3h 40m (6 steps)
Resources used: Helpers
--- Critical Path ---
Buy supplies -> Decorate venue -> Welcome guests -> Party games -> Serve cake -> Clean up
Float per step:
Buy supplies: 10:00 -> 11:00 [0m (critical)]
Bake cake: 10:00 -> 11:00 [95m slack]
Decorate venue: 11:00 -> 11:45 [0m (critical)]
Ice cake: 11:00 -> 11:20 [95m slack]
Welcome guests: 11:45 -> 11:55 [0m (critical)]
Blow up balloons: 11:00 -> 11:30 [15m slack]
Party games: 11:55 -> 12:55 [0m (critical)]
Set up games: 11:00 -> 11:20 [25m slack]
Serve cake: 12:55 -> 13:10 [0m (critical)]
Clean up: 13:10 -> 13:40 [0m (critical)]
--- Warnings ---
- Step 'Blow up balloons' was delayed beyond its available slack due to resource conflict with 'Helpers'
- Step 'Set up games' was delayed beyond its available slack due to resource conflict with 'Helpers'
Demonstrates: ALAP (As Late As Possible) timing for dinner steps, multi-day multi-track scheduling, and a long planning horizon across two days with no resource constraints.
$ npx skejj make examples/london-sightseeing.json
London Weekend Sightseeing
A two-day London itinerary covering major landmarks and cultural highlights. Dinner steps use ALAP timing to push them as late in the day as possible.
09:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:00
─ Day 1 ───────────────────────────────────────────────────────────────────────
Tower of London █████████│ │ │ │ │ │ │ │ │ │
Borough Market lunch │ │ █████│ │ │ │ │ │ │ │ │
Thames Southbank walk │ │ │ ████ │ │ │ │ │ │ │ │
Tate Modern │ │ │ │ ███████ │ │ │ │ │ │
Day 1 dinner │ │ │ │ │ │ ████ │ │ │ │ │
─ Day 2 ───────────────────────────────────────────────────────────────────────
British Museum │ │ │ │ │ │ │████████████ │ │ │
Covent Garden lunch │ │ │ │ │ │ │ │ │ ███ │ │
Westminster walk │ │ │ │ │ │ │ │ │ │ █████ │
London Eye │ │ │ │ │ │ │ │ │ │ │ ██│
Day 2 dinner │ │ │ │ │ │ │ │ │ │ │ █████
--- Summary ---
Total time: 12h
Steps: 10
Critical path: 12h (10 steps)
--- Critical Path ---
Tower of London -> Borough Market lunch -> Thames Southbank walk -> Tate Modern -> Day 1 dinner -> British Museum -> Covent Garden lunch -> Westminster walk -> London Eye -> Day 2 dinner
Float per step:
Tower of London: 09:00 -> 11:00 [0m (critical)]
Borough Market lunch: 11:00 -> 12:00 [0m (critical)]
Thames Southbank walk: 12:00 -> 12:45 [0m (critical)]
Tate Modern: 12:45 -> 14:15 [0m (critical)]
Day 1 dinner: 14:15 -> 15:15 [0m (critical)]
British Museum: 15:15 -> 17:45 [0m (critical)]
Covent Garden lunch: 17:45 -> 18:30 [0m (critical)]
Westminster walk: 18:30 -> 19:30 [0m (critical)]
London Eye: 19:30 -> 20:00 [0m (critical)]
Day 2 dinner: 20:00 -> 21:00 [0m (critical)]
See Schedule Template Schema Reference for the full field reference, or add a $schema key to your JSON file for editor autocomplete.
Schedules are JSON (or YAML) files. The fastest way to start:
# Interactive wizard -- no schema knowledge required
skejj new
# Or write JSON directly and validate as you go
skejj check myplan.jsonEditor autocomplete: Add a $schema field to your JSON file to get inline validation and autocomplete in VS Code and other editors:
{
"$schema": "./docs/schema.json",
"id": "my-schedule",
"name": "My Schedule",
"steps": [
{
"id": "step-1",
"title": "First step",
"durationMins": 30
}
]
}Full field reference: docs/SCHEMA.md
The schema covers: steps, tracks, resources (Equipment, People, Consumable), dependency types (FinishToStart, StartToStart, FinishToFinish, StartToFinish), timing policies (Asap, Alap), and time constraints (forward from startTime, backward from endTime).
