The Agent Skills Directory

[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it extracts and processes content from external PDF documents. Ingestion points: PDF files are processed using pypdf, pdfplumber, and pdftotext across SKILL.md and various scripts. Boundary markers: No specific delimiters or instructions (e.g., "ignore instructions in this text") are used to wrap extracted content. Capability inventory: The agent can execute shell commands via qpdf, pdftotext, and ImageMagick, and perform local file system writes. Sanitization: No sanitization or validation of the extracted PDF text is implemented before it enters the agent's context.\n- [SAFE]: The skill relies on well-established and trusted Python libraries, including pypdf, pdfplumber, and reportlab, for its primary operations.\n- [SAFE]: All command-line interactions (qpdf, pdftotext, magick) are standard for the intended workflow of local PDF manipulation and do not exhibit malicious patterns or privilege escalation attempts.\n- [SAFE]: No hardcoded credentials, sensitive file access patterns, or data exfiltration mechanisms were detected in the instructions or provided scripts.

pdf