The Agent Skills Directory

COMMAND_EXECUTION (HIGH): The script scripts/init-artifact.sh attempts to install pnpm globally using npm install -g pnpm if it is not found. This constitutes a privilege escalation attempt by modifying the system-wide environment and binaries.
EXTERNAL_DOWNLOADS (MEDIUM): The skill installs over 50 dependencies from the public npm registry during initialization and bundling. While these are common libraries (React, Tailwind, Radix UI), the sheer volume and the automated nature of the installation represent a significant supply chain risk.
REMOTE_CODE_EXECUTION (MEDIUM): The script scripts/init-artifact.sh uses node -e to execute dynamic JavaScript snippets for modifying tsconfig.json files. Additionally, the bundling process via Parcel involves executing complex build-time logic which can be exploited if project files are compromised.
PROMPT_INJECTION (LOW): The skill is vulnerable to indirect prompt injection (Category 8).
Ingestion points: The script scripts/init-artifact.sh accepts a user-controlled project name, and scripts/bundle-artifact.sh processes local files like index.html.
Boundary markers: None present to distinguish between instructions and data.
Capability inventory: The skill can execute shell commands, install packages, and perform file system operations.
Sanitization: No sanitization or validation is performed on the project name or the content of the files being bundled.

web-artifacts-builder