The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands, specifically python3 to run lab scripts (e.g., python3 phases/13-tools-and-protocols/06-mcp-fundamentals/code/main.py) and file system operations (rename, copy, delete) to manage learning state files like MCP-LEARNING.md and MCP-ENGINEERING-LEARNING.md.
[EXTERNAL_DOWNLOADS]: The skill fetches configuration files, documentation, and quiz data from the author's repository (https://raw.githubusercontent.com/fancyboi999/ai-engineering-from-scratch-zh/main/). This is a vendor-owned resource used to deliver the curriculum.
[INDIRECT_PROMPT_INJECTION]: The skill processes remote data from JSON and Markdown files to guide the tutoring session and command execution.
Ingestion points: learning-paths/model-context-protocol.json, docs/zh.md, and quiz.json (SKILL.md).
Boundary markers: No specific boundary markers or instructions to ignore embedded commands are specified for the ingested content.
Capability inventory: Subprocess execution (python3) and file system modifications (mv, cp, rm) across the tutoring workflow (SKILL.md).
Sanitization: The skill explicitly instructs the agent to redact secrets, tokens, and authorization headers before logging evidence to MCP-LEARNING.md (SKILL.md).
[DATA_EXFILTRATION]: The skill records the absolute working directory path into the local MCP-LEARNING.md file. While intended for checkpointing, logging system-specific paths constitutes metadata exposure.

learn-mcp