Skip to content

docker.md

Latest commit

 

History

History
109 lines (81 loc) · 3.3 KB

docker.md

File metadata and controls

109 lines (81 loc) · 3.3 KB
Raw

Docker

Run AtDork in a container — no local Python required. The image bundles the CLI, its dependencies and a baked copy of the Atdork-DB dork database, and runs as a non-root user (atdork).

Get the image

# published image (built & smoke-tested by CI on every push)
docker pull ghcr.io/amnottdevv/atdork:latest

# or build it from source
git clone https://github.com/amnottdevv/atdork.git
cd atdork
docker build -t atdork .

Version-pinned images are published for every release tag (ghcr.io/amnottdevv/atdork:1.4.1), plus latest on every push to master.

Quick start

alias atdork='docker run --rm -it -v "$PWD":/work ghcr.io/amnottdevv/atdork:latest'

atdork --version
atdork search "exposed s3 bucket" --limit 10
atdork --batch-file dorks.txt -r 20

/work is the container's working directory — mount your current directory so input files (--batch-file, --output) and results are visible on the host.

Dork database

The Atdork-DB checkout is baked into the image at /opt/atdork-db (selected via the ATDORK_DB_REPO environment variable). To keep installed packages and a fresh database between runs, add a named volume:

docker volume create atdork-data

docker run --rm -v atdork-data:/opt/atdork-db ghcr.io/amnottdevv/atdork:latest index
docker run --rm -v atdork-data:/opt/atdork-db ghcr.io/amnottdevv/atdork:latest install @cve/cve_1.txt
docker run --rm -v atdork-data:/opt/atdork-db ghcr.io/amnottdevv/atdork:latest installed

If the mounted directory is empty, the entrypoint clones the database automatically on first run. To use your own checkout instead:

docker run --rm -v "$HOME/Atdork-DB":/opt/atdork-db ghcr.io/amnottdevv/atdork:latest search cve

To refresh the baked database of a locally built image, just rebuild (docker build -t atdork .) — the clone step always fetches main.

Environment variables

Variable Default Purpose
ATDORK_DB_REPO /opt/atdork-db Root of the Atdork-DB checkout the package manager reads
ATDORK_DB_URL https://github.com/amnottdevv/Atdork-DB Repository cloned by the entrypoint when the database is missing

Recipes

Interactive shell inside the container

docker run --rm -it --entrypoint sh ghcr.io/amnottdevv/atdork:latest

GHDB scrape into a persistent store

docker run --rm -v "$PWD":/work -v atdork-data:/opt/atdork-db \
  ghcr.io/amnottdevv/atdork:latest --ghdb-scrape "exposed passwords" \
  --ghdb-after 2024-01-01 --output results.json

docker compose

services:
  atdork:
    image: ghcr.io/amnottdevv/atdork:latest
    volumes:
      - ./work:/work
      - atdork-data:/opt/atdork-db
volumes:
  atdork-data:
docker compose run --rm atdork search cve --limit 5

Image details

  • Base: python:3.12-slim (Debian), CLI installed with pip install .
  • Runtime user: atdork (non-root, uid auto) — mount with matching permissions for files you need to write on the host
  • git is included so atdork update and the first-run database bootstrap work
  • CI (.github/workflows/docker-publish.yml) builds the image and runs smoke tests (--version, index, search, install/installed, fresh-volume bootstrap) before publishing to GHCR on every push and tag