artifact-review

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a structured workflow for reviewing documentation artifacts using adversarial reasoning and specific checklists.
  • [COMMAND_EXECUTION]: Execution capabilities are restricted to authorized local project scripts (e.g., .claude/scripts/code_graph) and internal agent tools (spawn_agent). These are used for project indexing and analysis rather than malicious activity.
  • [SAFE]: No obfuscation, hidden URLs, or unauthorized network operations were detected. All file access is scoped to the local repository structure (e.g., docs/, .claude/).
  • [PROMPT_INJECTION]: The instructions include directives to ignore environment-specific mode switches (e.g., 'Ignore Claude-specific mode-switch instructions') and enforce strict execution contracts. These serve as operational constraints for the agent's compatibility layer rather than malicious safety bypasses.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 03:12 PM
Security Audit — agent-trust-hub — artifact-review