The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes external, untrusted PDF files and converts them to text that may be processed by an AI agent.
Ingestion points: The convert_pdf_to_markdown.py script ingests user-provided PDF files via the pypdf library.
Boundary markers: The output Markdown file does not contain delimiters or warnings to treat the content as untrusted data.
Capability inventory: The skill has the ability to write files to the local file system and communicate with the Mistral API via network requests.
Sanitization: The script does not perform sanitization or content filtering on the OCR results before writing them to disk.
[EXTERNAL_DOWNLOADS]: The script communicates with api.mistral.ai to perform OCR tasks. This is a well-known service used for the skill's primary functionality.
[COMMAND_EXECUTION]: Documentation in SKILL.md and references/reference.md describes the use of uv run and subprocess.run for script execution and batch processing tasks. These are documented for developer automation purposes.
[DYNAMIC_EXECUTION]: The scripts/test_image_namespacing.py script utilizes importlib.import_module to dynamically load the converter script. This is used exclusively for testing purposes and is a standard development pattern.

mistral-pdf-to-markdown