The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes user conversation history to generate documentation, which creates a surface for indirect prompt injection. However, since the primary purpose of the skill is documentation summarization, this is an inherent and expected behavior.
Ingestion points: Previous messages in the conversation context (SKILL.md).
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the guidelines.
Capability inventory: The agent performs directory listings (ls docs/), creates or writes files within the docs/ directory, and updates the AGENTS.md index file.
Sanitization: No explicit sanitization of the extracted conversation content is defined before the agent writes the data to the filesystem.
[COMMAND_EXECUTION]: The agent is instructed to run ls docs/ to identify current documentation categories. This is a standard read-only filesystem operation used for organization and does not represent a security risk.
[DATA_EXFILTRATION]: The skill provides an external link to the author's training platform (codely.com). This is a legitimate reference to the vendor's own domain and does not involve the transfer of sensitive data.

codely-doc-create