neta-adventure

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to utilize command-line tools for managing story campaigns, including creating, updating, and listing adventures. These tools are part of the developer's provided toolset.\n- [REMOTE_CODE_EXECUTION]: The skill uses npx -y @talesofai/neta-skills@latest to execute campaign management functions. This process involves downloading and running the latest version of the vendor's package from the npm registry at runtime. This is the documented mechanism for the agent to interface with the story backend.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-generated narrative content and loads external campaign or character data, which could potentially contain embedded instructions that influence the agent's behavior.\n
  • Ingestion points: Narrative data is loaded via the request_adventure_campaign and request_character_or_elementum tools in references/adventure-playing.md.\n
  • Boundary markers: The skill documentation recommends using the [OOC:] (Out-of-Character) prefix to distinguish user instructions from story dialogue.\n
  • Capability inventory: The skill has the ability to write to the story database using create_adventure_campaign and update_adventure_campaign.\n
  • Sanitization: No specific data sanitization or validation procedures for the story fields are described in the provided reference materials.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 09:53 AM