lionman888/us-crowding-radar

US Crowding Radar

Use the tested Python CLI in this repository. Do not reimplement scoring or call undocumented interfaces directly.

Safety

• This is read-only market intelligence, not trade execution.
• Never request, print, log, or place CLAWBY_API_KEY in command arguments.
• The dashboard accepts the key through its local UI, validates it against Clawby's official /api/account endpoint, and may save it to the Git-ignored local credential file with mode 600.
• CLI automation reads the key only from the environment or the ignored .env file.
• Never describe the crowding score as a price forecast or investment advice.

Commands

From the repository root:

python3 -m us_crowding_radar doctor
python3 -m us_crowding_radar watchlist list
python3 -m us_crowding_radar watchlist add TICKER --label "Company"
python3 -m us_crowding_radar watchlist remove TICKER
python3 -m us_crowding_radar scan
python3 -m us_crowding_radar scan --refresh
python3 -m us_crowding_radar scan --notify
python3 -m us_crowding_radar serve --port 8786

Workflow

1. For the dashboard, ask the user to enter their own Key in the local UI; never repeat or log it. A successful validation can be restored from the protected local credential file.
2. For CLI automation, run doctor and report only whether the environment key is configured.
3. Confirm the watchlist before a live scan.
4. Use scan for a cached authorized read-only run or scan --refresh for fresh remote data.
5. Report score, direction, coverage, component evidence, source timestamps, and failed dimensions together.
6. Only use --notify when the user has configured and authorized the notification target.

Validation

python3 -m unittest discover -s tests -v
node --check us_crowding_radar/web/app.js
python3 -m compileall -q us_crowding_radar