krt-light-compound-master
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from repositories and user requests, creating a surface for indirect prompt injection.
- Ingestion points: The skill is instructed to inspect the repository, current changes, user requests, and relevant requirements (SKILL.md).
- Boundary markers: There are no explicit instructions to use delimiters or ignore instructions embedded within the processed content.
- Capability inventory: The skill manages implementation (file system writes), verification (shell command execution), and release management (Git and Jira interactions).
- Sanitization: The instructions do not specify any sanitization or escaping of external data before processing.
- [COMMAND_EXECUTION]: The skill executes shell commands during its verification phase.
- Evidence: The execution and verification section explicitly mentions running "focused checks" and recording the "exact command" and its result in the state record (SKILL.md).
Audit Metadata