Setup guide

This is / isn’t

Use only content you have the right to use. Do not redistribute paid newsletter text. GuruWatcher arms concrete prices printed in the issue’s text. If a letter names a 200-day MA without a number, it may resolve that MA from the market — it does not invent a random level. Alerts are not trade instructions.

First hour — Control Panel (no terminal needed)

The Control Panel is a small local page that runs on this computer only. Dry-run is safe — nothing is sent to Discord until you go live. The packaged Mac/Windows app needs no Python install — Python is only needed for the source-zip path (macOS has python3 built in; Windows: install Python 3.10+ from python.org).

In one sentence: your newsletter arrives in email → GuruWatcher reads the prices → Discord pings when a level is hit. It never places trades.

Two safe modes, one difference: Run once (safe) = one price check right now, does not post to Discord. Quiet = Keep checking on a timer, still does not post until you choose Live with a saved webhook.

1. Open the Control Panel (this is the app)

   Packaged app (no Python needed): macOS — open the DMG, drag GuruWatcher to Applications, and launch it (notarized Developer ID; if Mac still blocks it: right-click → Open) · Windows — unzip the download and run GuruWatcher.exe (Authenticode-signed via Azure Artifact Signing; SmartScreen may still warn once for a new publisher — if so: More info → Run anyway).

   First time on Windows? The Keep → Unblock → Run anyway prompts are normal: Windows first run, step by step.

   Source zip: unzip the download. Double-click: macOS scripts/Open-GuruWatcher.command (if Mac says it can’t open: right-click → Open → Open) · Windows scripts/Open-GuruWatcher.bat.

   First run may take a few minutes. Your browser opens a local page (usually http://127.0.0.1:8788/). Packaged app: a tray icon keeps it running — reopen the panel or quit from the tray. Source launcher: leave the black terminal window open — that is the engine. Closing it turns the panel off.

2. Tell it which Discord channel to buzz (optional for practice)

   In Discord: open your channel → gear (Edit channel) → Integrations → Webhooks → New webhook → Copy URL. In the panel: paste under Discord alert link (your Discord webhook URL) → Save → Send test ping. You should see a short test message. Skip this step if you only want to practice with no Discord yet.

   Step-by-step: Connect a Discord webhook.

3. Connect your newsletter email (so you don’t paste every issue)

   What it does / doesn’t do: each check only pulls new or unread mail in one folder — not your whole inbox history. Use a dedicated mailbox or a Gmail filter into one label/folder — not your everyday personal inbox.

   In the panel → Newsletter email inbox: click Use Gmail, enter that mailbox’s address, paste an App password (Google Account → Security → App passwords), set the folder, and Save. Don’t want email yet? Open One-time paste → Load sample → Run once (safe) just to see it work.

   Step-by-step: Connect your newsletter email.

4. Turn on Keep checking

   Scroll to Keep checking. Defaults are fine: prices ~5 min, email ~15 min (separate timers). Start with Quiet if you want no Discord posts yet. Click Start checking. Your computer must stay awake.

   When you’re happy: choose Live (needs the webhook) → Start again. Discord only fires when a price level is actually hit — not for every email.

That’s the whole first hour. Power users can use the command line, Docker, YAML tuning, or MCP (below).

Advanced — CLI, Docker, scheduling, YAML

unzip GuruWatcher-v0.1.8.zip
cd GuruWatcher-v0.1.8

python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
# Or: pip install -e .

py -3 -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -r requirements.txt
# If Activate.ps1 is blocked:
#   Set-ExecutionPolicy -Scope Process Bypass

python3 -m guru_watcher init --wizard --data-dir ./gw-data   # interactive; --wizard --yes = headless
export GURU_WATCHER_DATA="$(pwd)/gw-data"
python3 -m guru_watcher status

cp "$GURU_WATCHER_DATA/.env.example" "$GURU_WATCHER_DATA/.env"
# Edit .env — only when you want live Discord:
# GURU_WATCHER_DISCORD_WEBHOOK_URL=https://discord.com/api/webhooks/...

mkdir -p "$GURU_WATCHER_DATA/drop"
# Drop a .md/.txt issue you own rights to into drop/
python3 -m guru_watcher ingest-run --adapter file_drop --data-dir "$GURU_WATCHER_DATA" --path drop --once
python3 -m guru_watcher ingest-issue --inbox "$GURU_WATCHER_DATA/inbox" --guru example --dry-run
python3 -m guru_watcher check --dry-run

export GURU_WATCHER_HOST_DATA="$(pwd)/gw-data"
export GURU_WATCHER_MODE=dry-run
docker compose up --build

python3 -m guru_watcher serve --data-dir "$GURU_WATCHER_DATA" --interval 900
python3 -m guru_watcher serve --once --data-dir "$GURU_WATCHER_DATA"   # single pass

pip install -e '.[mcp]'          # from a checkout
# or:  pip install 'guru-watcher[mcp]'

{
  "mcpServers": {
    "guru-watcher": {
      "command": "python3",
      "args": ["-m", "guru_watcher", "mcp"],
      "env": { "GURU_WATCHER_DATA": "/absolute/path/to/gw-data" }
    }
  }
}

Scheduling

ingest-run is a single pass — schedule it yourself. Docker loop-pipeline covers the full spine on an interval, and serve does the same without Docker.

Non-Docker loop — serve

python3 -m guru_watcher serve --data-dir "$GURU_WATCHER_DATA" --interval 900
python3 -m guru_watcher serve --once --data-dir "$GURU_WATCHER_DATA"   # single pass

Runs ingest → reconcile → check on an interval. Dry-run default; --mode live requires a webhook and fails closed.

Linux / macOS cron

*/15 * * * * cd /path/to/GuruWatcher && GURU_WATCHER_DATA="$HOME/gw-data" .venv/bin/python -m guru_watcher ingest-run --adapter file_drop --data-dir "$HOME/gw-data" --path drop --once
*/5  * * * * cd /path/to/GuruWatcher && GURU_WATCHER_DATA="$HOME/gw-data" .venv/bin/python -m guru_watcher check --dry-run

Windows Task Scheduler

$py = "$env:USERPROFILE\GuruWatcher\.venv\Scripts\python.exe"
$data = "$env:USERPROFILE\gw-data"
$cmd = "& '$py' -m guru_watcher check --dry-run"
$action = New-ScheduledTaskAction -Execute "powershell.exe" -Argument "-NoProfile -ExecutionPolicy Bypass -Command $cmd"
$trigger = New-ScheduledTaskTrigger -Once -At (Get-Date) -RepetitionInterval (New-TimeSpan -Minutes 5)
Register-ScheduledTask -TaskName "GuruWatcherCheck" -Action $action -Trigger $trigger

No Docker? Linux templates in systemd/toolkit/ run the same pipeline via a venv.

Support

• Primary: Support page (support@guruwatcher.com)
• After you have the toolkit (repo invite available on request), you can also open a GitHub Issue.
• Community (optional): Discord.

Platform matrix

docs/PLATFORM.md: pip/venv CLI on Linux/macOS/Windows; Docker pipeline on Linux + Docker Desktop; CI smoke on Linux (host+docker), macOS (pytest), Windows (host ps1). Packaged Control Panel app (PyInstaller — no Tauri) for macOS Apple Silicon (DMG) and Windows (portable zip); no native GUI redesign.

Safety

Alert-only BYO sources No orders

• Never places trades
• Do not commit webhook URLs or API keys
• Not investment advice · not a broker · not redistributing paid newsletters