The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input provided in the system description to categorize risks and generate compliance reports. * Ingestion points: User-provided [系统名称或AI用例描述] (System name or description) is extracted and used in Step 3 of the workflow to determine regulatory roles and risk levels. * Boundary markers: The skill lacks explicit delimiters or instructions to ignore potential commands embedded within the user-provided descriptions. * Capability inventory: The skill has capabilities to read from and write to the filesystem at ~/.claude/plugins/config/claude-for-legal-zh/ai-governance-legal/. * Sanitization: There is no evidence of sanitization or validation of the input strings before they are used to influence the logic of the assessment output.

aia-generation