code-review-axes-and-quality
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because its core function involves reading and analyzing untrusted source code. Malicious instructions embedded in comments or strings within the code under review could attempt to influence the agent's analysis or final verdict.
- Ingestion points: The
PROJECT_CONTEXTvariable inscripts/review-plan.template.jsand the instructions inSKILL.mdspecify that the agent reads complete files and diffs from the target project. - Boundary markers: The templates for sub-agent prompts do not include specific delimiters or instructions to ignore potential commands found within the code blocks being analyzed.
- Capability inventory: The system employs
subagentcalls and possesses the capability to read and write files within the.agents/plans/directory of the workspace. - Sanitization: The skill does not implement sanitization or escaping of the ingested code content before it is interpolated into the worker prompts.
- [COMMAND_EXECUTION]: The workflow requires the generation and execution of JavaScript files (based on the
scripts/review-plan.template.jstemplate) within a code execution block. This execution is necessary to orchestrate the sequential or parallel review axes, involving the dynamic handling of project-specific paths and context.
Audit Metadata