Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Cloud Docker Api Abuse

DSecurity

Turn an exposed container daemon or an over-privileged container into host compromise — a Docker API on 2375/2376, a mounted docker.sock, or a privileged/`-v /:/host` container. Load when you find an open Docker/containerd socket, port 2375/2376, a container you can run images in, or you landed inside a container and want the host. Signals: 2375/tcp open, /var/run/docker.sock reachable, `docker` group membership, privileged container, host paths mounted in.

20 stars
0 votes
0 copies
0 views
Added 9/22/2026
ai-agentsgodockerapisecurity

Works with

cliapi

Security Analysis

D42/100
mediumUses curl or wget to download content
criticalAccesses sensitive system or user directories
criticalAccesses sensitive system or user directories

Pro shows the line behind each finding and how to fix it

Scanned 9/22/2026

$npx -y skills add NoorQureshi/SploitAgent --skill cloud-docker-api-abuse --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Cloud Docker Api Abuse?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Cloud Docker Api Abuse
[![Security: D — Skills Directory](https://www.skillsdirectory.com/api/skills/noorqureshi-cloud-docker-api-abuse/badge)](https://www.skillsdirectory.com/skills/noorqureshi-cloud-docker-api-abuse)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: cloud-docker-api-abuse
description: >
  Turn an exposed container daemon or an over-privileged container into host compromise — a Docker
  API on 2375/2376, a mounted docker.sock, or a privileged/`-v /:/host` container. Load when you
  find an open Docker/containerd socket, port 2375/2376, a container you can run images in, or you
  landed inside a container and want the host. Signals: 2375/tcp open, /var/run/docker.sock reachable,
  `docker` group membership, privileged container, host paths mounted in.
domain: cloud
type: technique
stability: learning
modes: [pentest, bugbounty]
severity: critical
owasp: [A05:2021-Security-Misconfiguration]
mitre: [T1610, T1611]
cwe: [CWE-16, CWE-284]
tools: [docker, curl, netexec]
schema_version: 1
---

# Docker / container daemon abuse

## When it applies
You can reach a container control plane you shouldn't: an unauthenticated Docker API on `2375/tcp`
(or `2376` without client-cert auth), a mounted `/var/run/docker.sock` inside a container you
control, membership in the `docker` group, or a privileged/host-mounted container. Any of these is
effectively root on the host. Complements `cloud-container-escape` (which focuses on kernel/runtime
breakouts); this one is about the daemon/socket as the vector.

## Why it works
The Docker daemon runs as root and will build/run any container with any mount you ask for. Anyone
who can talk to its API or socket can start a container that bind-mounts the host filesystem
(`-v /:/host`) or runs `--privileged` — then read/write host files, add a root user, or chroot in.
There's no privilege boundary between "can use the daemon" and "root on the host".

## Method
1. **Confirm the vector.**
   - Remote API: `curl http://<host>:2375/version` and `/containers/json` — a response = unauthenticated control.
   - In-container socket: `ls -la /var/run/docker.sock`; `docker` CLI or `curl --unix-socket`.
   - `docker` group: `id | grep docker`.
2. **Point the CLI at the daemon** (remote): `export DOCKER_HOST=tcp://<host>:2375` then `docker ps`.
3. **Mount the host and take root** — run a container bind-mounting the host root and act on it:
   ```
   docker run -v /:/host --rm -it alpine chroot /host sh
   ```
   From there: add a privileged user, drop an SSH key into `/root/.ssh/authorized_keys`, read
   `/etc/shadow`, or (Windows host) pull `NTDS`/SAM.
4. **Privileged-container escape** (already inside one): `--privileged` exposes host devices —
   mount the host disk (`fdisk -l` → `mount /dev/sdaN /mnt`) or abuse cgroups release_agent to run
   a command as root on the host.
5. **Pull secrets** from the daemon: image layers, env vars (`docker inspect`), and other containers'
   filesystems often hold credentials → pivot with `recon`/`cloud-*`.

## Gotchas
- **`2376` is usually TLS + client-cert** — if it rejects you without a cert, it's *not* the open one;
  `2375` is the plaintext/no-auth port to look for.
- **Read-only or restricted socket mounts** may allow listing but not `run` — check what verbs work
  before assuming full control.
- **Non-destructive by default** — creating containers/users changes host state; on bounty confirm
  RCE minimally (e.g. read a root-only file) and stop, per `reporting-triage-validation`.
- **Clean up** every container, image, and user you create.

## Verify success
Read or write a host-only resource from your mounted/escaped context (e.g. `/etc/shadow`, a file in
another user's home, or a root-owned path) — proving host-level access from the container plane.

## References
Docker Engine API docs; `docker.sock` escape write-ups; MITRE ATT&CK T1610/T1611. Kernel/runtime
breakouts: `cloud-container-escape`.

Attribution

NoorQureshiNoorQureshi
View sourceSee grades on GitHubMore from NoorQureshi →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698461 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →