gt-create-issue
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the local codebase and user instructions to generate issue titles and bodies, creating a potential vector for indirect prompt injection if codebase files contain malicious instructions.
- Ingestion points: User instructions and content from the local codebase are explored and used to draft issue content (SKILL.md).
- Boundary markers: No explicit delimiters or warnings to ignore embedded instructions within the codebase are provided.
- Capability inventory: The skill has the capability to execute the
gtCLI tool to create issues on a remote repository. - Sanitization: There is no evidence of sanitization or escaping of the ingested content before it is used in the issue creation command.
- [COMMAND_EXECUTION]: The skill executes the
gtCLI tool to perform actions on a Forgejo repository. - Evidence: The instruction
gt issue create -R <owner/repo> --title "<title>" --body-file <body-file>(SKILL.md) directly invokes a shell command. - Mitigation: The skill includes a human-in-the-loop checkpoint in step 2, requiring the agent to show the draft and ask the user for confirmation before proceeding to create the issue.
Audit Metadata