The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's core script, scripts/process.py, automatically detects the absence of the Pillow library and installs it from the Python Package Index (PyPI) to ensure the tool functions correctly.
[COMMAND_EXECUTION]: To handle dependency management, the skill executes system-level commands using the subprocess module to invoke the pip package manager.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of image metadata, which could contain malicious instructions designed to influence the agent.
Ingestion points: The scripts/process.py script reads image files from the file system, which may contain attacker-controlled metadata tags.
Boundary markers: The script's output does not use delimiters or explicit 'ignore instructions' warnings to isolate metadata content from the rest of the output provided to the agent.
Capability inventory: The skill possesses file read/write capabilities and the ability to execute subprocesses for installation tasks.
Sanitization: While the script truncates long metadata strings, it does not filter or sanitize the actual content of EXIF tags for potential instructional commands before they are displayed.

image-tools