惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
U
Unit 42
Google DeepMind News
Google DeepMind News
博客园 - 司徒正美
Y
Y Combinator Blog
F
Fortinet All Blogs
云风的 BLOG
云风的 BLOG
T
Tailwind CSS Blog
G
Google Developers Blog
酷 壳 – CoolShell
酷 壳 – CoolShell
罗磊的独立博客
D
DataBreaches.Net
T
The Blog of Author Tim Ferriss
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
MyScale Blog
MyScale Blog
N
Netflix TechBlog - Medium
Microsoft Security Blog
Microsoft Security Blog
GbyAI
GbyAI
P
Proofpoint News Feed
Jina AI
Jina AI
B
Blog RSS Feed
腾讯CDC
阮一峰的网络日志
阮一峰的网络日志
D
Docker

Hacker News: Show HN

PurrrrrFocus: Pomodoro Timer App - App Store Workflow Engine — Multi-Step Orchestration for Bun RapidPhoto: Pro Photo Editor App - App Store GitHub - DheerG/swarms: Achieve extraordinary results with claude code across a variety of tasks SPICE simulation → oscilloscope → verification with Claude Code — Lucas Gerads Show HN: VCoding – A 5 MB native Windows IDE with no dynamic dependencies Show HN: LLMs don't hallucinate because they're bad at math, it's the format GitHub - Agent-FM/agentfm-core: AgentFM is a peer-to-peer network that turns everyday computers into a decentralized AI supercomputer. AgentFM lets you run massive AI workloads directly across a global mesh of idle CPUs and GPUs. Show HN: Tracking Top US Science Olympiad Alumni over Last 25 Years GitHub - Potarix/agent-hub: One place to talk to all your agents Show HN: Runtime security for AI agents(injection,tool abuse, data exfiltration) GitHub - dubeyKartikay/lazyspotify: Terminal Spotify client for macOS and Linux GitHub - the-banana-tool/king-louie: Easy to use GUI Personal AI Assistant. Win/Linux/Mac. Show HN I made my vacation rental bookable by AI agents–no Airbnb, 0% commission GitHub - basteez/jsf-autoreload: maven plugin to enable hot reload on jsf projects uvm32/hosts/host-gdbstub at main · ringtailsoftware/uvm32 GitHub - labsai/EDDI: Config-driven engine that turns JSON into production-grade AI agents. Multi-agent orchestration, 12+ LLM providers, MCP/A2A protocols, RAG, persistent memory, and enterprise compliance (EU AI Act, GDPR, HIPAA). Built on Quarkus. GitHub - glitchnsec/fortyone-oss: AI Executive Assistant Platform Quickstart | Alien GitHub - muxshed/shed: One stream in, or many. Every destination, simultaneously. No cloud middleman, no per-channel fees, no limits. GitHub - ocrbase-hq/ocrbase: 📄 PDF/IMG ->.MD/JSON Document OCR API for PaddleOCR and GLMOCR. Self-hostable. GitHub - impactjo/home-memory: MCP server that lets your AI assistant remember everything about your home. GitHub - Sets88/dbcls: DbCls is a powerful terminal database client that supports various databases GitHub - neptun2000/heor-agent-mcp GitHub - SeanFDZ/macmind: Single-layer transformer in HyperTalk for the classic Macintosh RollQuation: Math Puzzles - Apps on Google Play GitHub - dropbox/witchcraft Show HN: Agent-cache – Multi-tier LLM/tool/session caching for Valkey and Redis GitHub - opentalon/opentalon: OpenTalon is an open-source platform built from the ground up in Go as a robust alternative to OpenClaw LinkedIn™ 职位抓取工具 - Chrome 应用商店
GitHub - anmalkov/image-inspector: 🐳 Find and digest-pin ...
anmalkov · 2026-06-26 · via Hacker News: Show HN

🐳 image-inspector — select • inspect • pin

image-inspector

A CLI for finding official container base images, showing precomputed vulnerability counts, and generating digest-pinned FROM lines — without pulling images, running Docker, or scanning locally.

CI status Latest release PyPI version MIT License


⚡ Try it in 5 seconds

No install, no Docker daemon, no local scanner — just uv:

uvx --from base-image-inspector image-inspector

Pick a base image with the arrow keys and copy the digest-pinned FROM line. That's it.

image-inspector result panel showing a digest-pinned FROM line and vulnerability counts

Why this exists

Tool Great at The gap it leaves
docker pull + trivy scan Accurate, thorough scanning Slower, and runs locally — you pull the image first
Renovate Keeping base images up to date Helps after you've already chosen a base image
image-inspector Choose + inspect + pin before you write FROM Approximate counts from precomputed nightly data (fetched from GitHub Pages, bundled offline fallback), not a live scan

Who is this for?

Use image-inspector if you:

  • write Dockerfiles often
  • want reproducible base images
  • want quick vulnerability context before choosing a base image
  • don't want to pull images or run a scanner locally

What is this?

When you write a Dockerfile, you start from a base image like python:3.13 or node:22. The problem: tags like python:3.13 are moving targets — the image behind that tag changes over time. So a build that works today might pull a different image tomorrow, and "it works on my machine" quietly breaks.

image-inspector fixes that. You pick a language or OS, a version, and a variant — all with the arrow keys — and it gives you a base image pinned to an immutable digest plus a ready-to-paste FROM line:

FROM python:3.13.14-slim@sha256:205e60d0b78f024817...

It also shows you, up front, how many known security vulnerabilities that image has, its size, and when it was built — so you can choose a good base image with confidence.

What's that @sha256:... part? It's the image's digest — a unique fingerprint of the exact image contents. Pinning to a digest means everyone who builds your Dockerfile gets the identical base image, every time. That's what makes a build reproducible.

Vulnerability counts come from precomputed nightly Trivy data — fetched from GitHub Pages when online, with a copy bundled in the package as an offline fallback. Images are not pulled or scanned locally at runtime. No Docker daemon or local scanner is required.

image-inspector demo

Quick start

1. Already tried it with uvx? Install it permanently (pick whichever you have):

uv tool install base-image-inspector     # recommended
# or
pipx install base-image-inspector
# or
pip install base-image-inspector

Package vs. command: the PyPI package is base-image-inspector, but the installed CLI command is image-inspector.

Prefer one-shot usage? Use uv:

uvx --from base-image-inspector image-inspector

2. Run it:

3. Pick with the arrow keys — language/OS → version → variant — and copy the FROM line it prints. That's it. 🎉

New here and want the full walkthrough? See the Getting started guide.

Features

  • 📌 Digest pinning — outputs a name:tag@sha256:… reference for reproducible builds.
  • 🛡️ Security at a glance — critical / high / total vulnerability counts for the chosen image, from precomputed nightly Trivy data fetched from GitHub Pages (with a bundled offline fallback).
  • 🧱 Many ecosystems, one interface — Python, .NET, Java, Go, Node, Rust, C/C++, plus Ubuntu, Debian and Alpine base images.
  • 🤖 Automation-friendly--json for non-interactive use and --plain / NO_COLOR support.
  • 🎨 Modern UI — branded banner, themed menus, spinners, and a syntax-highlighted result panel.
  • ⌨️ Arrow-key everything — language, version, and variant are all pick-from-list menus. No typing.
  • 📋 Quick actions — after a result, copy the FROM line or digest to your clipboard.

Supported images

Languages & runtimes

Language Registry Repository Versioning
Python Docker Hub library/python semver (latest 5 minors)
.NET MCR mcr.microsoft.com/dotnet/sdk semver (latest 5 minors)
Java Docker Hub library/eclipse-temurin feature release (8 / 11 / 17 / 21 / 25 / 26)
Go Docker Hub library/golang semver (latest minors)
Node.js Docker Hub library/node semver (latest 5 minors)
Rust Docker Hub library/rust semver (latest 5 minors)
C / C++ Docker Hub library/gcc semver (latest 5 minors)

OS base images

Image Registry Repository Versioning
Ubuntu Docker Hub library/ubuntu calver YY.MM (latest 5 releases, LTS marked)
Debian Docker Hub library/debian major (11 / 12 / 13) + -slim variant
Alpine Docker Hub library/alpine semver (latest 5 minors)

Per-image details (Java feature releases, the gcc compiler image, Ubuntu LTS, Debian variants) are covered in the Getting started guide.

Examples

# Interactive — pick everything with the arrow keys:
image-inspector

# Non-interactive, machine-readable output for scripts/CI:
image-inspector --json -l ubuntu --version 24.04

A --json run prints a single object describing the resolved image. For example:

image-inspector --json -l python --version 3.13 --variant slim
{
  "source": "Docker Hub",
  "language": "python",
  "version": "3.13",
  "variant": "slim",
  "image": "python:3.13.14-slim",
  "pinned_reference": "python:3.13.14-slim@sha256:205e60d0b78f024817...",
  "digest": "sha256:205e60d0b78f024817...",
  "size_bytes": 44912345,
  "from_line": "FROM python:3.13.14-slim@sha256:205e60d0b78f024817...",
  "vulnerabilities": {
    "critical": 0,
    "high": 1,
    "total": 23,
    "scanned_at": "2026-06-22T02:14:07+00:00"
  },
  "scanner": { "name": "trivy", "version": "0.71.1", "db_updated_at": "2026-06-22T00:00:00+00:00" }
}

(Some fields are omitted above for brevity.) When no scan data exists for the image, vulnerabilities is null.

The full list of flags lives in the Getting started guide.

Vulnerability data

The critical / high / total counts come from precomputed nightly Trivy data. Nothing is scanned locally at runtime — image-inspector doesn't run Trivy on your machine, pull images, or talk to a scanner. That keeps it fast and means no Docker daemon or scanner is required. A GitHub Actions workflow regenerates this data nightly and publishes it to GitHub Pages. At runtime the tool is online-first: it fetches that live report when online (short timeout, ETag-cached) and falls back to the copy bundled with the package when offline or if the fetch fails. The SECURITY panel's Source row shows which you're seeing (online (latest) vs offline (bundled copy)). Set IMAGE_INSPECTOR_OFFLINE=1 to force the bundled copy, or IMAGE_INSPECTOR_REPORT_URL to point at a different report. Because the data is precomputed, counts reflect the most recent snapshot rather than a live, on-the-spot scan.

Limitations

  • Vulnerability counts come from the precomputed nightly dataset (online from GitHub Pages, or the bundled offline fallback), not a live scan.
  • Counts are for the selected base image only, not your final application image.
  • Digest pinning improves reproducibility, but you still need a process for updating pinned images.
  • Only selected official images are supported.

Why not just use Trivy, Docker Scout, or Renovate?

image-inspector is not a replacement for full image scanning, Docker Scout, Trivy, or dependency automation tools like Renovate.

It is meant for the moment before you write a FROM line: choosing among official base images, seeing approximate vulnerability counts, and pinning the exact digest without pulling images locally or running a scanner.

You should still scan your final built image in CI.

Documentation

Community & support

🤝 Contributing

Contributions are welcome! Please read CONTRIBUTING.md for the branch/PR flow, local checks, and where to ask questions.

License

Released under the MIT License.