The Agent Skills Directory

[SAFE]: The skill consists entirely of natural language instructions guiding the AI's reasoning and output formatting for legal research. It does not contain any shell scripts, Python code, or executable binaries.
[PROMPT_INJECTION]: No attempts to bypass safety filters, ignore system instructions, or extract system prompts were detected. The instructions include safety-positive directives, such as instructing the agent to stop if key legal bases are missing rather than using model memory to fill in gaps.
[DATA_EXPOSURE]: The skill does not access sensitive local file paths (e.g., SSH keys, AWS credentials) or environment variables. No hardcoded credentials were found.
[EXTERNAL_DOWNLOADS]: The author's homepage (chenshi.ai) is included in the metadata for reference. There are no patterns involving automated downloading and execution of remote scripts or packages.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to process user-provided legal facts and questions, it lacks the high-privilege capabilities (such as arbitrary command execution or file system writing) that would make it a significant security risk. The instructions focus on structured data processing and reporting.

issue-memo