The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the grep utility to perform keyword searches within files located in the AI_MEMORY/ directory. This is a legitimate use of command-line tools for the skill's stated purpose of information retrieval.
[INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for ingesting data from external files, which presents a surface for indirect prompt injection if the source data is attacker-controlled.
Ingestion points: Data is read from AI_MEMORY/INDEX.md and various files in level-specific subdirectories (e.g., AI_MEMORY/L1_情境层/, AI_MEMORY/L4_核心层/).
Boundary markers: No specific delimiters or safety warnings regarding the content of the memory files are present in the instructions.
Capability inventory: The skill is capable of executing shell commands (grep) and reading local files.
Sanitization: There are no instructions for sanitizing or validating the content retrieved from memory files before using it to generate responses.

mem-query