The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions to install the Coolify CLI by piping remote scripts directly to the shell (curl | bash and irm | iex). This pattern is high risk as it executes unverified code from an external repository.
Evidence in SKILL.md: curl -fsSL https://raw.githubusercontent.com/coollabsio/coolify-cli/main/scripts/install.sh | bash
Evidence in SKILL.md: irm https://raw.githubusercontent.com/coollabsio/coolify-cli/main/scripts/install.ps1 | iex
[DATA_EXFILTRATION]: The skill instructions direct the agent to access and upload sensitive local files, specifically private SSH keys, to the remote service.
Evidence in SKILL.md and references/command-reference.md: coolify private-key add --name deploy-key --file ~/.ssh/id_ed25519
[COMMAND_EXECUTION]: The skill is designed around executing arbitrary shell commands via the Coolify CLI to manage infrastructure, which provides a broad attack surface for server manipulation.
Evidence: Extensive use of the coolify command across all instruction files.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes external .env files which could contain malicious instructions designed to influence the agent's behavior during synchronization tasks.
Ingestion points: Local .env files via coolify app env sync --file .env (referenced in references/command-reference.md).
Boundary markers: None mentioned.
Capability inventory: Subprocess calls via CLI, file system access, and network operations.
Sanitization: No sanitization or validation of the file content is described.

coolify-cli