The Agent Skills Directory

[SAFE]: The skill implements a robust 'goal contract' system designed to prevent agent overreach and unsafe autonomous actions by requiring explicit boundaries, stop conditions, and pause conditions.
[SAFE]: The Python script scripts/lint_goal_contract.py is a utility for validating text files; it performs regex checks for required markers (Verification, Constraints, etc.) and flags placeholders or vague/dangerous language. It does not perform network operations, access sensitive files, or execute external code.
[SAFE]: All referenced files are local to the skill directory. No external dependencies or remote execution patterns were found.

goal-setter