fable5-prompting-guide
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of informational markdown files and prompt templates intended to guide model behavior. It does not contain any executable scripts, binaries, or automated shell commands.\n- [SAFE]: The instructions include specific safety-conscious guidance, such as advising against attempting to bypass model safeguards and providing methods to handle refusals as API states rather than attempting to prompt around them.\n- [SAFE]: The skill promotes the use of structured XML tags to delineate untrusted inputs and external documents. This practice is recommended to prevent ambiguity and mitigate potential indirect prompt injection in the agents that will use these prompts.\n- [SAFE]: References to external resources point exclusively to official and trusted domains associated with well-known technology organizations. These links provide context for model runtime configurations and best practices without introducing external risks.
Audit Metadata