The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill describes a workflow for managing infrastructure based on user requirements. It follows best practices by providing a dedicated secret management system to handle sensitive inputs. (1) Ingestion points: User-defined infrastructure configurations and requirements across various providers (e.g., Cloudflare, AWS, Neon) defined in alchemy.run.ts files. (2) Boundary markers: The instructions use standard code blocks and hierarchical project structures to organize code, following common developer environment conventions. (3) Capability inventory: The skill enables full infrastructure lifecycle management via the alchemy CLI, involving network requests to provider APIs and local file system modifications. (4) Sanitization: The framework includes a built-in alchemy.secret() method for secure handling of sensitive data and uses physical naming conventions to ensure resource isolation.
[DYNAMIC_EXECUTION]: The framework facilitates the generation and execution of TypeScript configuration files to manage cloud resources, which is a core and expected function for Infrastructure-as-Code tools.
[SAFE]: The skill focuses on providing legitimate developer documentation and reference material. It does not exhibit malicious intent, obfuscation, or unauthorized access patterns, and it correctly manages sensitive data paths for its own internal operations.

alchemy