lcs-code-review
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from the local repository to perform code reviews. \n
- Ingestion points: The skill reads project source code, diff files, and LCS artifacts such as
explore.md,prd.md, andsrs.mdfrom the project workspace. \n - Boundary markers: The instructions lack explicit requirements for using boundary markers or protective prompt wrappers to prevent the agent from obeying instructions that might be embedded within the code or artifacts being reviewed. \n
- Capability inventory: The skill has permissions to read repository files and write report artifacts to the
.lcs/work-items/directory, which could be misused if an injection attack successfully influences the agent's behavior. \n - Sanitization: There are no documented procedures for sanitizing, escaping, or filtering the content of code files or documentation artifacts before they are processed by the LLM.
Audit Metadata