cmd-gh-issue
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill generates shell commands using the
ghCLI (gh repo viewandgh issue create). To prevent shell expansion and command injection from the synthesized conversation, it correctly uses a quoted heredoc (<<'ISSUE_EOF') for the issue body. The--titleparameter is also encapsulated in double quotes, though users should ensure the agent-generated title doesn't contain breaking characters. - [DATA_EXFILTRATION]: By design, the skill extracts conversation content and transmits it to GitHub's servers to create an issue. While this is the intended functionality, it constitutes a data flow where potentially sensitive information from the chat history (e.g., environment details or internal logs) is moved to an external platform.
- [PROMPT_INJECTION]: The skill is subject to indirect prompt injection (Category 8).
- Ingestion points: Conversation history (referenced in instructions).
- Boundary markers: None explicitly defined to separate data from instructions within the synthesis step.
- Capability inventory: Execution of
ghCLI commands (SKILL.md). - Sanitization: Employs quoted heredocs to safely handle the transition of text into the shell command body.
Audit Metadata