The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted image data through OCR, creating a surface for indirect prompt injection where text within images could be interpreted as instructions.
Ingestion points: Image files such as input.png mentioned in SKILL.md are processed by the tesseract tool.
Boundary markers: No delimiters or specific instructions are provided to separate the extracted OCR text from agent instructions.
Capability inventory: The skill performs local CLI execution of tesseract and rg (ripgrep).
Sanitization: SKILL.md identifies basic Unicode and whitespace normalization in the post-processing phase.
[EXTERNAL_DOWNLOADS]: The skill references documentation and code from official Tesseract OCR repositories.
Evidence: URL references to https://github.com/tesseract-ocr/tessdoc and https://github.com/tesseract-ocr/tesseract located in references/tesseract-ocr-2026-02-18.md.

tesseract-ocr