add-backend

Warn

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill provides templates (templates/python_backend.py) and instructions (SKILL.md) that utilize dynamic module loading via importlib.import_module. This mechanism is used to lazily load and execute backend logic based on a registry of module paths, which is a common but sensitive pattern for plugin-based architectures.
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of new data ingestion points that process external arguments (**kwargs) without providing built-in sanitization or boundary markers in the generated templates.
  • Ingestion points: Untrusted data enters the agent context through CLI arguments passed to the invoke function and subsequently to the backend's execute method.
  • Boundary markers: The provided Python and Rust backend templates do not include delimiters or instructions to ignore embedded commands in the processed data.
  • Capability inventory: Backends created using this skill have the capability to execute arbitrary Rust native code or Python module logic, as evidenced in the lib.rs and templates/rust_backend.py files.
  • Sanitization: There is no evidence of input validation, escaping, or filtering in the provided templates; the skill leaves the implementation of security guardrails entirely to the developer.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 14, 2026, 03:58 AM