aomi-build
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill demonstrates a high level of security awareness by declaring restricted permissions and following least-privilege principles.
- [COMMAND_EXECUTION]: The skill uses local 'cargo' and 'git' binaries to manage Rust projects. This execution is limited to specific allowed subcommands and is essential for its primary function of scaffolding and building code within the user's workspace.
- [DATA_EXFILTRATION]: The skill is configured with a strict no-network policy (network.deny: "*"). This ensures that no data can be sent to external destinations by the skill itself, and instructions specify avoiding embedding credentials in generated code.
- [PROMPT_INJECTION]: The skill provides an interface for processing untrusted external data like OpenAPI specifications (ingestion point: SKILL.md). While this creates an attack surface for indirect prompt injection, the skill mitigates this by enforcing strict architectural patterns for generated code (boundary: instructions in references/spec-to-tools.md), using restricted shell and write capabilities, and explicitly denying the creation of build.rs files (sanitization: deny_write in SKILL.md).
Audit Metadata