The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell scripts (scripts/init-artifact.sh and scripts/bundle-artifact.sh) to initialize the environment and bundle the final output.
[EXTERNAL_DOWNLOADS]: The bundling process involves downloading and installing several Node.js packages from public registries, including parcel, @parcel/config-default, parcel-resolver-tspaths, and html-inline.
[PROMPT_INJECTION]: The skill operates by transforming user-provided descriptions into functional code artifacts, creating a surface for indirect prompt injection. Maliciously crafted user prompts could attempt to influence the agent's behavior or embed instructions into the generated codebase.
Ingestion points: User requests and design specifications provided to the agent for artifact creation (referenced in the general development workflow).
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the skill definition.
Capability inventory: Includes shell command execution (bash), file system writes (editing generated files), and network access for package installation.
Sanitization: The skill does not define specific sanitization or validation logic for the input used to generate code.

artifacts-builder