gh-templates
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local repository files to customize scaffolded templates, creating an attack surface for indirect prompt injection.
- Ingestion points: The agent is instructed to read repository files such as
package.json,Makefile, and.github/workflows/*.yml(as specified inSKILL.md) to extract project-specific lint, test, and build commands. - Boundary markers: The instructions do not define boundary markers or delimiters for the content read from these files before it is interpolated into template bodies.
- Capability inventory: The skill performs file system write operations to create templates in the
.github/directory and can execute shell commands via theghCLI. - Sanitization: There is no evidence of sanitization or validation for the strings extracted from repository files before they are written into markdown templates.
- [COMMAND_EXECUTION]: The skill enables the agent to execute GitHub CLI commands to modify repository settings.
- The agent is instructed to offer to create repository labels using
gh label createfor bug, enhancement, and chore types. - Evidence: The skill explicitly includes a safeguard requiring user confirmation before executing these commands to prevent unauthorized changes to repository configuration.
Audit Metadata