The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches and installs the Helvstack CLI utility from the developer's official repository at github.com/ElyzeSolutions/Helvstack.
[COMMAND_EXECUTION]: Utilizes system shell commands to perform repository analysis, application building, and deployment operations via the helvstack CLI.
[CREDENTIALS_UNSAFE]: The skill manages sensitive data by interacting with .env files and local configuration paths (.helvstack/config.json). It implements strong security controls, including the use of umask 077 and chmod 600 for file protection, and explicit instructions to never log or version-control secrets.
[DATA_EXFILTRATION]: Performs necessary network operations to the vendor's console and API endpoints (console.helvstack.com) for agent authentication and resource provisioning.
[PROMPT_INJECTION]: The skill ingests untrusted data from the user's repository files while maintaining administrative capabilities.
Ingestion points: Reads project documentation (e.g., README.md, AGENTS.md) and package manifests to model the deployment.
Boundary markers: Not explicitly utilized for file reading; however, the skill mandates machine-readable JSON output and strict output filtering.
Capability inventory: Possesses full command execution, file system access, and network operation capabilities.
Sanitization: Includes strict directives against reading unscoped secrets and forbids echoing sensitive values in any output or terminal state.

helvstack-agent-deploy