implement
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch external data from an issue tracker, including ticket titles and content for the primary task and related sibling tickets. This creates a surface for indirect prompt injection where an attacker could place malicious instructions inside a ticket to manipulate the agent.
- Ingestion points: External issue tracker and related ticket references (SKILL.md).
- Boundary markers: No delimiters or specific warnings to ignore embedded instructions are specified for the fetched ticket content.
- Capability inventory: Shell command execution for typechecking, test suite runs, and git commits (SKILL.md).
- Sanitization: The instructions only require verifying the success of the fetch operation (using
[NEED-VERIFY]) but do not mention sanitizing or validating the content of the tickets themselves. - [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for regular typechecking, running test files, executing full test suites, and committing code to a repository. While these are legitimate actions for the skill's purpose, they provide the necessary capabilities for an injection attack to cause unauthorized system changes.
Audit Metadata