The Agent Skills Directory

[SAFE]: The skill is designed as a template-driven instructional guide for verifying business data and logic.
[DATA_EXPOSURE]: The skill attempts to read ../../CLAUDE.md for project-level configuration. In the context of developer-oriented AI agents, this is a standard practice for retrieving environment rules and does not target sensitive OS-level files or user credentials.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data (insight lists and business backgrounds). While it lacks explicit boundary markers (e.g., XML tags) to isolate this data, the skill possesses no exploitable capabilities such as network access, file writing, or command execution, rendering the risk negligible.
[METADATA]: The skill metadata and versioning are consistent with its stated purpose of insight validation in a financial or business context.

insight-validation