Skip to content

Competition Container Runtime

Diagnoses how live containers and pods differ from their manifests in CTF sandbox work

Stars
43,969
Licence
MIT
Files
3 files

What it does

This is a specialist add-on for a CTF (capture-the-flag security competition) sandbox workflow. When a live container or Kubernetes pod behaves differently from its checked-in config, the agent is guided to compare the manifest against what is actually running: mounted secrets, sidecars, init containers, entrypoints, and the route from a hostname through a reverse proxy to the right container. It is meant to be used only after a parent skill, ctf-sandbox-orchestrator, has already set up the sandbox assumptions, and it asks the agent to reply in Simplified Chinese by default.

Great for

  • CTF players or competition organisers debugging a containerised challenge environment
  • Developers tracing why a deployed container's runtime state differs from its compose or Kubernetes manifest
  • Students learning how mounts, sidecars and reverse-proxy routing interact at runtime

Skip it if

  • Production systems or anything outside an authorised sandbox, since the skill focuses on finding exploitable runtime paths and secrets, which is risky if misapplied
  • Users who have not installed the parent ctf-sandbox-orchestrator skill, because the file says this skill should not be used on its own

Try asking

  • “Why does the live container's config differ from the docker-compose file I checked in?”
  • “Where is this mounted secret actually read by a process inside the pod?”
  • “Which route on my reverse proxy resolves to which running container?”
  • “How does this init container change the runtime state of the app?”
Step-by-step guide: how to use Competition Container RuntimeInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    competition-container-runtime.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip competition-container-runtime.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip competition-container-runtime.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Install and activate ctf-sandbox-orchestrator first, otherwise the agent is told to route back to it.
  • Provide the manifest, the live container or pod ID, and the relevant mount paths so the agent can compare them.
  • Ask in English if you need English output, since the default reply language is Simplified Chinese.

What's inside

  • agents/openai.yaml
  • references/container-runtime.md
  • SKILL.md

Redistributed unmodified from zhaoxuya520/reverse-skill under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.