The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local Python script (scripts/generate_differentiated.py) that processes teacher inputs and generates Word documents using the python-docx library.
[PROMPT_INJECTION]: User-provided inputs such as the subject, lesson title, grade, and student needs are interpolated directly into a shell command string and a file output path. This creates an attack surface for indirect prompt injection, command injection, and path traversal if the inputs contain malicious shell characters or path sequences. Mandatory Evidence Chain (Category 8): 1. Ingestion points: User inputs collected during Step 1 in SKILL.md. 2. Boundary markers: Absent in the bash command template. 3. Capability inventory: Bash tool for script execution and Write tool for document generation. 4. Sanitization: None visible in the provided instructions.
[SAFE]: The skill reads several context files from parent directories (../../tw_edu_...) and absolute paths (/mnt/skills/public/docx/SKILL.md). These appear to be part of the vendor's (FW1201) shared educational framework and are documented as necessary configuration.

tw-edu-differentiated