The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to download a remote repository from https://github.com/buildermethods/design-os.git. This external source is not affiliated with the stated author of the skill and originates from an unverified third party.- [REMOTE_CODE_EXECUTION]: The documentation guides the user to perform a git clone followed by npm install and npm run dev. This sequence facilitates the execution of arbitrary remote code from an untrusted external repository.- [COMMAND_EXECUTION]: Multiple shell commands are utilized to set up the environment and run the application, including git clone, npm install, bun install, npm run dev, npm run build, and npm start.- [INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill ingests and processes untrusted data defining product entities, section specifications, and feature requirements provided by the user or defined in the skill documentation.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard potential instructions embedded within the data models or product specifications.
Capability inventory: The skill environment supports shell command execution (npm, git, bun) and outbound network requests to AI service providers.
Sanitization: The documentation does not specify any sanitization or validation mechanisms to filter malicious content from the design specifications before they are processed by the tool or the AI.

design-os-product-planning