ralph-tui-create-github-issues
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute various
gh(GitHub CLI) commands to create issues, view issue details, and interact with the GitHub GraphQL API. - [COMMAND_EXECUTION]: The skill generates a temporary GraphQL mutation file at
/tmp/blocked_by.graphqland executes it via thegh api graphqlcommand to avoid shell encoding issues. - [PROMPT_INJECTION]: The skill processes untrusted PRD content and interpolates it into shell command bodies, creating a surface for indirect prompt injection.
- Ingestion points: PRD markdown or text files provided as input (processed in SKILL.md).
- Boundary markers: The skill explicitly mandates the use of single-quoted HEREDOC delimiters (
<<'EOF') to prevent the shell from interpreting content inside the issue body. - Capability inventory: GitHub Issue creation, viewing, and GraphQL API interaction via the
ghCLI (described in SKILL.md). - Sanitization: The instructions provide specific warnings and examples on how to use HEREDOC syntax safely to prevent variable expansion and command substitution.
Audit Metadata