PushPress CLI

A small CLI that automates the PushPress Members web app to:

• Log in
• Extract workouts
• Book classes by day/time

This tool is built for the client/member experience at https://members.pushpress.com.

Requirements

• Node.js
• Playwright browsers

Install dependencies:

npm install

Install Playwright browsers:

npx playwright install

Quick Start

1. Create a .env file (see Configuration below).
2. Install dependencies and Playwright browsers.
3. Run list to see available flows.
4. Run a flow, optionally with --no-headless --pause --verbose for debugging.

Example:

npx tsx src/cli.ts list
npx tsx src/cli.ts run login --no-headless --pause --verbose

Configuration

Create a .env file in the project root (or pass --config <path>):

PUSHPRESS_EMAIL=you@example.com
PUSHPRESS_PASSWORD=your-password
PUSHPRESS_BASE_URL=https://members.pushpress.com
OPENAI_API_KEY=your-openai-key

Optional OpenAI settings (used for workout-week markdown summaries):

• OPENAI_MODEL: default gpt-3.5-turbo-16k
• OPENAI_PROMPT_PATH: any file path (default: ./prompts/workout-week-summary.md)

Example with a custom config file:

npx tsx src/cli.ts config --config ./envs/pushpress.env

Validate the config before running a flow:

npx tsx src/cli.ts config --config ./envs/pushpress.env --validate

Commands

List available flows:

npx tsx src/cli.ts list

Show resolved config:

npx tsx src/cli.ts config

Validate config:

npx tsx src/cli.ts config --validate

Validate current session:

npx tsx src/cli.ts validate-session --verbose

Run a flow:

npx tsx src/cli.ts run <flow-name>

Global Options

All known values:

• --config <path>: any file path (default: .env)
• --verbose: boolean flag
• --headless / --no-headless: boolean flag (default: --headless)
• --slow-mo <ms>: integer milliseconds (e.g., 250)
• --timeout <ms>: integer milliseconds (e.g., 30000)
• --pause: boolean flag
• --dry-run: boolean flag (logs steps without executing)

Common option patterns:

# Debug a flow with a visible browser and slower actions
npx tsx src/cli.ts run workout-week --no-headless --slow-mo 250 --pause --verbose

# Use a non-default config file
npx tsx src/cli.ts run login --config ./envs/pushpress.env

# Increase timeouts for slow networks
npx tsx src/cli.ts run workout-history --timeout 60000 --verbose

Flows

login

Logs into the Members app and saves a session to state/session.json.

npx tsx src/cli.ts run login --no-headless --verbose

workout-history

Extracts workout history (network capture) while on the Workouts tab.

npx tsx src/cli.ts run workout-history --no-headless --verbose

Optional:

• --workout-type <name>: selects a workout type from the dropdown (e.g., Bootcamp, HYROX; quote multi-word names like "PRVN Pump").
  • When provided, output filenames include the workout type slug (e.g., workout-history-201530-bootcamp.json).

Typical usage:

npx tsx src/cli.ts run workout-history --no-headless --verbose --pause

workout-week

Clicks each day in the week and captures workouts for all days. Output includes weekly arrays.

npx tsx src/cli.ts run workout-week --no-headless --verbose --pause

Optional:

• --workout-type <name>: selects a workout type from the dropdown (e.g., PRVN Burn; quote multi-word names like "PRVN Pump").
  • When provided, output filenames include the workout type slug (e.g., workout-week-201530-prvn-burn.json).

This flow also writes a summary file grouped by day with only title, description, and workoutTitle, plus the detected date:

output/workout-week/YYYY-MM-DD/workout-week-HHmmss-summary.json
output/workout-week/YYYY-MM-DD/workout-week-HHmmss-<workout-type>-summary.json

If OPENAI_API_KEY is set, it also writes a formatted markdown summary using the prompt in ./prompts/workout-week-summary.md:

output/workout-week/YYYY-MM-DD/workout-week-HHmmss-summary.md
output/workout-week/YYYY-MM-DD/workout-week-HHmmss-<workout-type>-summary.md

Example: JSON only (no OpenAI key):

npx tsx src/cli.ts run workout-week --no-headless --verbose

Example: with OpenAI markdown summary:

OPENAI_API_KEY=your-openai-key npx tsx src/cli.ts run workout-week --no-headless --verbose

schedule-book

Books classes for specific days and a time. This flow is dry-run by default.

Required:

• --days <list>
• --time <label>

Optional: All known values:

• --days <list>: comma/space separated day keys: sun, mon, tue, wed, thu, fri, sat
• --time <label>: time label as shown in UI, e.g. 6:00 AM, 5:00 PM
• --class <name>: class name label as shown in UI (default: CrossFit; quote multi-word names like "Olympic Lifting").
• --type <name>: alias for --class (quote multi-word names).
• --category <name>: one of Reservations, Classes, Appointments, Events (default: Classes)
• --week <which>: current, next, or an integer offset like 2, 3 (max 6)
• --waitlist: allow waitlist-only booking
• --confirm: perform booking actions (without this it is dry-run)

Example dry-run:

npx tsx src/cli.ts run schedule-book --days mon,wed,fri --time "5:00 PM" --class "CrossFit" --no-headless --verbose --pause

Example confirm booking:

npx tsx src/cli.ts run schedule-book --days mon,wed,fri --time "5:00 PM" --class "CrossFit" --confirm --no-headless --verbose --pause

Target next week:

npx tsx src/cli.ts run schedule-book --week next --days fri --time "5:00 PM" --class "CrossFit" --confirm --no-headless --verbose --pause

Target next-next week:

npx tsx src/cli.ts run schedule-book --week 2 --days fri --time "5:00 PM" --class "CrossFit" --confirm --no-headless --verbose --pause

Common booking variations:

# Book a different class name and allow waitlist
npx tsx src/cli.ts run schedule-book --days tue,thu --time "6:00 AM" --class "Olympic Lifting" --waitlist --confirm

# Book from the Reservations category
npx tsx src/cli.ts run schedule-book --category Reservations --days mon --time "12:00 PM" --confirm

# Dry-run for a future week offset
npx tsx src/cli.ts run schedule-book --week 3 --days fri --time "5:00 PM"

Output Files

All flows (except login) write JSON output to:

output/<flow-name>/<YYYY-MM-DD>/<flow-name-HHmmss>.json

For schedule-book, the output includes:

• data.bookings for confirmed reservations
• data.matches for dry-run matches
• data.attempts for skipped or attempted bookings

Troubleshooting

• If login fails, delete state/session.json and run login again.
• If the UI is slow or flaky, use --timeout 60000 and --slow-mo 250.
• If the flow seems stuck, rerun with --no-headless --pause --verbose to inspect the browser state.
• If OpenAI summaries fail, verify OPENAI_API_KEY and OPENAI_PROMPT_PATH (if set).

Notes

• The CLI will reuse state/session.json when available. If the session is invalid, it re-authenticates.
• Use --pause to keep the browser open at the end of a run for debugging.