The Agent Skills Directory

[SAFE]: The skill is designed strictly for requirements gathering and documentation. It includes explicit constraints (ソースコード・実装ファイル・設定ファイルの編集・作成・削除は禁止) that prevent the agent from modifying the project's source code or configuration.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user input and web search results (現状調査の実行, Web 検索). While this presents a surface for indirect prompt injection where a malicious web page could attempt to influence the requirement definition, the impact is low because the skill's output is limited to a Markdown document and it does not have the capability to execute code or change system state.
[DATA_EXPOSURE]: The skill performs research by reading project files and directory structures. This is a standard and necessary function for requirements definition and does not involve exfiltrating sensitive credentials or personal data to unauthorized third parties.

define