The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the makecli binary from the author's repository via Homebrew and includes an update mechanism via makecli update.
[COMMAND_EXECUTION]: Shell commands are utilized for OCR operations, configuration verification, and credential management. It also documents the use of the !command syntax for interactive token setup.
[DATA_EXFILTRATION]: User-provided document files are uploaded to the configured service URL for recognition and processing.
[PROMPT_INJECTION]: The skill processes untrusted external files, creating an indirect prompt injection surface.
Ingestion points: Extracts data from user-provided .pdf, .ofd, .png, .jpg, and .jpeg files using makecli (SKILL.md).
Boundary markers: Lacks explicit instructions for the agent to ignore instructions embedded within the invoice/bill content during extraction.
Capability inventory: Includes shell command execution and recording operations via makecli (SKILL.md).
Sanitization: No explicit sanitization or validation of extracted OCR results is performed before they are processed by the agent.

make-integration