The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves pull request data from GitHub, including titles, descriptions, and comments. This externally-sourced content represents an indirect prompt injection surface as it could contain malicious instructions designed to manipulate the agent's behavior when the data is processed.
[DYNAMIC_EXECUTION]: The query-prs.sh script uses eval to execute a command string constructed from several input parameters such as --repo, --author, and --search. These variables are not sanitized before being concatenated into the command string, which introduces a command injection vulnerability if the agent is coerced into providing malicious arguments.

github-pr-query