making-code-agent-friendly
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes a JavaScript script
shared/scripts/contention.mjsthat executes localgitcommands (log,rev-parse,ls-files) to perform contention analysis. These operations are performed usingnode:child_process.spawnwith argument arrays, which is a secure method for executing CLI tools that avoids shell injection vulnerabilities. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted repository data, such as git commit messages, PR comments, and file contents. It mitigates the risk of indirect prompt injection through a comprehensive security policy in
shared/external-data.md, which mandates that all such external data be treated strictly as data and never as instructions. The policy also includes technical safeguards against shell interpolation. - [SAFE]: The skill demonstrates excellent security hygiene. It instructs the agent to validate untrusted file names against a strict character allowlist (
^[A-Za-z0-9._/@+-]+$) before using them. Furthermore, it enforces the use of fresh-context, read-only subagents for code reviews to maintain independence and prevent context leakage.
Audit Metadata