ai-agents-architect
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill provides architectural documentation and best practices for building AI agents. It does not contain any executable scripts, shell commands, or automated tasks. All code snippets provided are for instructional purposes (pseudocode) rather than direct execution.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns or package installation commands were detected. The skill does not perform any downloads or execute external scripts.
- [DATA_EXFILTRATION]: There are no network operations, API calls, or sensitive file system accesses. The skill does not handle or transmit data.
- [PROMPT_INJECTION]: The instructions focus on system design and do not contain any attempts to override model behavior, bypass safety filters, or extract system prompts.
- [OBFUSCATION]: No obfuscated content, encoded strings, or hidden characters were found in the skill metadata or body.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or instructions to store credentials unsafely were identified.
Audit Metadata