The Agent Skills Directory

[SAFE]: The skill consists entirely of high-level instructional content, logical frameworks, and output templates. No malicious patterns or behaviors were identified during the analysis.\n- [NO_CODE]: No functional code blocks, shell scripts, or binary files are present in the skill definition. The skill relies solely on the model's natural language reasoning capabilities.\n- [PROMPT_INJECTION]: The instructions are focused on ensuring high-quality legal analysis and do not attempt to override the AI's safety guidelines or system instructions. While the skill processes user-provided legal facts (an ingestion point in SKILL.md), it lacks any dangerous capabilities (subprocess, network, or file writing) that could be exploited via indirect prompt injection.\n- [DATA_EXFILTRATION]: No network access commands (e.g., curl, wget) or attempts to access sensitive system files (e.g., .ssh, .env) were detected.\n- [COMMAND_EXECUTION]: There are no shell commands or patterns indicating the execution of arbitrary system code, including no instances of dynamic context injection using the '!command' syntax.

structured-element-extraction