The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes arbitrary academic papers from untrusted sources such as user-provided URLs or PDF uploads.
Ingestion points: External academic papers provided by users via URL or PDF attachment (SKILL.md, Step 1).
Boundary markers: None detected. The instructions do not define specific delimiters or provide warnings to the agent to disregard instructions potentially embedded within the paper text.
Capability inventory: The agent utilizes shell commands (pdftotext), the file tool (read/write/view), and the message tool to process and deliver data.
Sanitization: No sanitization, filtering, or validation of the extracted text is performed before the content is used for analysis in Step 2 and report generation in Step 3.
[COMMAND_EXECUTION]: The skill instructs the agent to execute the pdftotext command-line utility to extract content from PDF files, which involves shell-level interaction with user-supplied data (SKILL.md, Step 1).
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch PDF documents from arbitrary user-provided URLs (SKILL.md, Step 1).

paper-parse