apostle-artifacts-loops
Pass
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses sophisticated instructional tags (e.g.,
<identity>,<zero-trust>,<governing>) to define agent behavior. There are no attempts to bypass safety filters or override system instructions for malicious purposes. - [DATA_EXFILTRATION]: No network operations or unauthorized file access patterns were detected. The skill manages project-specific markdown files (Plan.md, Memo.md, etc.) within the local workspace to maintain state.
- [REMOTE_CODE_EXECUTION]: The skill does not contain any commands for downloading or executing external scripts. It relies on standard markdown file operations.
- [COMMAND_EXECUTION]: No shell commands or direct process executions are requested. The logic is handled through the agent's internal reasoning and file system interactions.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or instructions to access sensitive credential stores were found.
- [OBFUSCATION]: The skill mentions using token-efficient formats for agent-to-agent communication, such as 'Lean', 'Grug-speak', or 'Classical Chinese'. These are described as optimization techniques for long-context tasks rather than attempts to hide malicious logic.
Audit Metadata