OpenParable
Skip to content
← Back to skills

Hackthebox

Apps & Automation
hackthebox

Hackthebox is an Agent Skill for apps & automation published in the transilienceai/communitytools repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

HackTheBox platform operations and automations to solve challenges, machines and capture the flags hacking competitions

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/transilienceai/communitytools/blob/HEAD/skills/hackthebox/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/hackthebox, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Workflow

  • workflow.md — Complete workflow with commands. Read this for each step

Steps

  1. Get Credentials — python3 .claude/tools/env-reader.py HTB_USER HTB_PASS HTB_TOKEN ANTHROPIC_API_KEY SLACK_BOT_TOKEN HTB_SLACK_CHANNEL_ID
  2. Only for "Machine" kind of competition -> Verify vpn is running, otherwise download the vpn file from the platform and instruct the user on how to enable it
  3. Generate output dirs — mkdir -p YYMMDD_<name>/{recon,findings,logs,artifacts,tools,reports} for each challenge. Do not seed attack-chain.md or experiments.md from the orchestrator — those are the coordinator subagent's first action (see skills/coordination/reference/spawning-recipes.md).
  4. To achieve the tasks given by the user, when possible use the HTB_TOKEN, otherwise login to the platform using playwright at https://account.hackthebox.com/login and fill the login form with the HTB_USER and HTB_PASS
  5. If necessary, start the machines
  6. If necessary, check network connectivity to the machines
  7. Spawn coordinator subagent per target — Agent(name="coordinator-{tag}", run_in_background=True, ...) per spawning-recipes.md. Never run the P0-P6 coordinator workflow inline in the orchestrator session — the bookkeeping discipline (attack-chain.md, experiments.md, goal_attempts counting, mandatory skeptic at experiments 5/15/25) requires the subagent boundary. Max N concurrent agents, queue-based spawning.
  8. Post-solve Phase 3 — parent orchestrator (not coordinator) always runs /skill-update + Slack after each coordinator returns its PHASE3_SUMMARY (see workflow.md step 8)

References