The Agent Skills Directory

[COMMAND_EXECUTION]: The generatePdfCurl function in generate.js is vulnerable to shell command injection. It constructs a curl command string by interpolating the outputPath variable directly into a string that is then executed via child_process.execSync. A maliciously crafted filename could escape the single quotes used in the command string to execute arbitrary shell commands on the host system.
[DATA_EXFILTRATION]: The skill transmits the content of reports to an external third-party API (api.nutrient.io) to perform HTML-to-PDF conversion. While this behavior is disclosed in the skill's documentation, it involves sending potentially sensitive user data to a remote service, which represents a data exfiltration risk if the documents contain private information.
[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by reading and processing markdown files from the filesystem. If these files are provided by an untrusted source, they could contain instructions designed to influence the behavior of the agent during the generation process.
Ingestion points: Markdown input files read using fs.readFileSync in generate.js.
Boundary markers: No boundary markers or instructions are provided to the agent to treat document content as passive data.
Capability inventory: The skill has the ability to execute shell commands (execSync), write files to the disk (fs.writeFileSync), and perform network operations (axios, curl).
Sanitization: The markdown-to-HTML parser in generate.js performs only minimal HTML escaping for code blocks and lacks comprehensive sanitization for the rest of the content.

elegant-reports