depheal 0.1.0

dependency health scanner — cross-language, offline-first, no account needed

depwise

dependency health scanner — no account, no api key, no config

$ depwise scan ~/netbox-test

depwise — dependency health scanner

  dir      ~/netbox-test
  reading  requirements.txt, pyproject.toml

  scanning 45 packages...

  Django@6.0.5   high  5 CVEs  fix: 5.2.15
    An issue was discovered in Django 6.0 before 6.0.6 and 5.2 before 5.2.15
    PYSEC-2026-200, PYSEC-2026-198 +3 more

  colorama@0.4.6   abandoned
    No updates in 3 years

  django-graphiql-debug-toolbar@0.2.0   abandoned
    No updates in 4 years

  1 vulnerable, 2 abandoned, 42 ok

  to fix:
    pip install Django==5.2.15

that's a real scan of NetBox — used in production by NVIDIA, Cloudflare, and thousands of others.

the problem with existing tools

pip-audit and npm audit exist. but:

• they audit the wrong environment when you're inside a venv
• they show 40 CVEs with no context — developers learn to ignore them
• they don't know if a package is abandoned (no CVE required to be dangerous)
• they're single-language — mixed projects need multiple tools

depwise fixes all of this.

install

pip install depwise

no account. no api key. no config file. works immediately.

usage

# scan current directory
depwise

# scan any directory from anywhere
depwise scan ./myproject
depwise scan ~/anyproject

# explain a specific package
depwise why requests
depwise why flask --version 2.2.0

# list all packages found
depwise list

# use in CI/CD — exits with code 1 if issues found
depwise scan --strict

works with:

• requirements.txt
• pyproject.toml
• package.json

what makes it different

detects abandoned packages — a package with no CVE but no maintainer is still a risk. depwise checks last commit dates and deprecation notices. existing tools don't.

right environment — automatically detects your active venv and scans that. pip-audit scans the wrong python when you're inside a venv.

one output — python and javascript in the same project, one scan, one report.

zero noise — shows what matters. one line per package. plain english.

zero dependencies — pure python stdlib. nothing to break. works everywhere python works.

how it works

• reads your dependency files
• detects your active virtual environment automatically
• queries OSV for known CVEs — free, no key needed
• checks PyPI and npm registry for abandoned/deprecated packages
• shows you what matters, not everything

license

MIT