The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary function involves the agent executing a local Node.js script (scripts/modao-capture.js) with user-supplied arguments via the bash tool. Executing local scripts can lead to arbitrary code execution on the host system.
[EXTERNAL_DOWNLOADS]: The environment setup requires running npm install, which downloads third-party packages from the NPM registry. Because the package.json file and its dependency list were not provided in the skill contents, these external dependencies cannot be verified for safety.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it scrapes content (page text and comments) from external Modao prototype links. Malicious instructions embedded in a prototype's metadata or comments could be processed by the agent in subsequent steps.
Ingestion points: External Modao prototype URLs (modao.cc) processed by the scripts/modao-capture.js script.
Boundary markers: No delimiters or instructions to ignore embedded commands are present in the provided documentation.
Capability inventory: Ability to execute Node.js scripts, perform network requests (via Puppeteer), and write files to the local project directory.
Sanitization: No evidence of input sanitization or safety filtering for the scraped external content before it is read by the agent.

modao-capture