dialogue-manager

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides a link to the official GitHub repository for the 'godot_dialogue_manager' addon (github.com/nathanhoad/godot_dialogue_manager). This is a standard reference for plugin installation.
  • [DYNAMIC_EXECUTION]: The documentation describes the create_resource_from_text(text) method. This function allows for the compilation of dialogue logic from a string at runtime, which is a core feature of the tool for procedural dialogue generation but involves dynamic interpretation of script-like content.
  • [INDIRECT_PROMPT_INJECTION]: The skill documents an attack surface where data processed by the Dialogue Manager (specifically .dialogue files) can trigger logic execution via 'mutations' and 'conditions'.
  • Ingestion points: External .dialogue resource files and raw strings passed to the create_resource_from_text function in SKILL.md (Section 4).
  • Boundary markers: The provided documentation does not specify the use of delimiters or warnings to prevent the engine from obeying instructions embedded within the dialogue text.
  • Capability inventory: The dialogue system can call methods on global objects (do SomeGlobal.method()) and modify global state variables (set SomeGlobal.property = true), which grants the dialogue content direct influence over game logic.
  • Sanitization: There is no mention of sanitization or validation logic for the input dialogue data within the skill's instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 07:09 AM