The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes Git shell commands (fetch, pull, stash, commit, push) to synchronize the ideas repository with a remote server.
[COMMAND_EXECUTION]: The skill attempts to execute a local Python script scripts/update_readme_ideas_index.py if present in the repository root to refresh the project index.
[DATA_EXFILTRATION]: The skill is designed to push idea descriptions and validation reports to a remote repository on GitHub (github.com/luongnv89/ideas). This involves transmitting user-provided content to an external service as part of its primary function.
[PROMPT_INJECTION]: The skill processes untrusted user input to populate documentation templates, creating an attack surface for indirect prompt injection (Category 8).
Ingestion points: Idea descriptions provided via $ARGUMENTS and user responses to clarification questions are interpolated into the idea.md file.
Boundary markers: No explicit delimiters or safety instructions are used to separate user-provided content from the template structure.
Capability inventory: The agent has the capability to write to the local file system, execute Git commands, and run local Python scripts.
Sanitization: There is no evidence of sanitization, escaping, or validation of the user input before it is written to markdown files.

idea-validator