The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to install several Python libraries using pip if they are missing: python-docx, reportlab, matplotlib, japanize-matplotlib, and pypdf. These are standard, well-known libraries for document generation and analysis.
[COMMAND_EXECUTION]: The skill uses the curl command to upload the final document to an external endpoint provided within the task's context. This is a functional requirement for document delivery.
[DATA_EXFILTRATION]: The final step of the process involves sending the generated document to a remote server via a PUT request. While this involves sending data externally, it is the primary intended function of the skill for delivering the generated report.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from web research and user-provided inputs without explicit sanitization or boundary markers.
Ingestion points: External data sources and web research content cited in the document (SKILL.md).
Boundary markers: None identified; instructions do not specify delimiters for external content.
Capability inventory: The skill has the ability to write files to the local system (python-docx, reportlab) and perform network operations (curl) (SKILL.md).
Sanitization: No specific sanitization or filtering of the ingested external content is mentioned.

professional-document