The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches shell scripts and auxiliary skill metadata from the author's GitHub repository at github.com/Melvynx/ai-maxxing-skills. These downloads are used to facilitate the SSH configuration process.
[REMOTE_CODE_EXECUTION]: The skill downloads and executes a shell script (ssh-setup.sh) locally to manage SSH keys and configuration. This script is hosted on the author's public repository.
[COMMAND_EXECUTION]: The skill executes several shell commands to manage the system's SSH state, including ssh-keygen for key generation and ssh for remote connectivity tests. It also executes remote commands on the VPS to collect system metadata such as OS details, hardware specifications, and the presence of Docker or Dokploy.
[PERSISTENCE]: The skill makes several persistent changes to the local environment: it modifies ~/.ssh/config to save connection aliases, creates rule files in ~/.claude/rules/vps.md to inform future agent sessions of the server's existence, and updates ~/.claude/settings.json to automatically allow specific SSH, SCP, and Rsync commands.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources (VPS provider panels) via browser snapshots. This untrusted content is used to identify IP addresses and login credentials, which are subsequently used in local configuration scripts and permission settings.
Ingestion points: Browser snapshots of VPS provider panels (Hostinger, Hetzner, etc.) in SKILL.md step 4.
Boundary markers: None identified.
Capability inventory: The skill can write to the file system (~/.ssh/config, ~/.claude/settings.json), execute local shell scripts, and initiate network connections via SSH.
Sanitization: The skill prompts the user to verify the IP address and take a screenshot before proceeding with critical steps.

connect-vps