The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external documents (.md, .txt, .pdf) for analysis and summarization. This represents a potential surface for indirect prompt injection; however, the skill lacks high-privilege capabilities such as network access, shell command execution, or sensitive file writing that could be leveraged for exploitation. The risk is effectively neutralized by the absence of dangerous tools.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns, package installations, or external script downloads were detected.
[DATA_EXFILTRATION]: The skill does not perform network operations or access sensitive environment variables or credentials. It operates strictly on provided educational materials.
[OBFUSCATION]: No obfuscated content, encoded strings, or hidden instructions were found in the skill metadata or body.

learning-methodology