The Agent Skills Directory

[SAFE]: The skill operates locally by reading and writing files within a specific, scoped directory (~/.claude/plugins/config/claude-for-legal-zh/). No unauthorized file access or network communication was detected.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from a configuration file and user-provided system descriptions to generate reports. While this creates a minor surface for processing untrusted text, the skill is purely informational and does not trigger automated commands or sensitive actions based on the content, mitigating the risk of exploitation.
[DATA_EXPOSURE]: The instructions handle organizational AI inventory and compliance data. All data processing is confined to the local environment and the skill does not contain hardcoded credentials or mechanisms to exfiltrate data to external servers.

ai-inventory