github-goal-issue-triage
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which creates a surface for indirect prompt injection.
- Ingestion points: The agent fetches
GOAL.mdfiles and GitHub issue content (titles, bodies, and comments) to determine priority levels and next-step nominations. - Boundary markers: The instructions do not specify the use of clear delimiters or instructions to ignore embedded commands when the agent processes the goal or issue text.
- Capability inventory: The skill possesses the ability to create and modify GitHub labels and write an HTML file to the local filesystem.
- Sanitization: While the skill requires HTML escaping for the generated report to prevent XSS, it lacks sanitization for the content interpolated into the agent's decision-making context.
- [COMMAND_EXECUTION]: The skill instructions allow the agent to use the
ghCLI tool to perform repository operations. While this is necessary for the skill's primary function, it grants the agent the ability to execute commands that could be targeted if the agent is influenced by malicious input from a repository's issues or goal files.
Audit Metadata