The Agent Skills Directory

[DYNAMIC_EXECUTION]: The test suite in scripts/test_decision_calculator.py uses subprocess.run to invoke the decision_calculator.py script for unit testing. This is a standard and safe practice within a development environment, as it uses hardcoded paths and the current Python interpreter to verify local script behavior without exposing the system to arbitrary command injection.
[INDIRECT_PROMPT_INJECTION]: The decision_calculator.py utility processes user-supplied JSON decision models. The script implements defensive programming by performing strict schema validation, verifying probability sums, and enforcing finite numerical types, effectively preventing schema confusion or data-driven injection attacks through the tool's input surface.
[SAFE]: The skill's content, including mathematical playbooks and academic references, aligns entirely with its stated purpose of providing foundations for decision theory. There are no signs of credential harvesting, hidden exfiltration chains, or persistence mechanisms in the provided instructions or source code.

foundations-decision-theory