Update to RUMBO Agent Reliability

{
  "description": "Keep agent work aligned to one measurable goal with explicit gates and stop conditions.",
  "included_files": [],
  "name": "goal-loop-controller",
  "skill_md_contents": "---\nname: goal-loop-controller\ndescription: Keep agent work aligned to one measurable goal with explicit gates and stop conditions.\n---\n\n# Goal Loop Controller\n\nRun: GOAL -> STATE -> OPTIONS -> NEXT ACTION -> VERIFY -> RECONCILE -> GATE.\n\nFor each iteration:\n1. State one concrete goal.\n2. Read current state before acting.\n3. Pick the smallest action that can materially advance the goal.\n4. Stay inside authorized scope.\n5. Verify independently when possible.\n6. Update state and evidence.\n7. Stop when the next action needs missing authority, credentials, external approval, or unproven evidence.\n\nDo not repeat unchanged diagnostics, create parallel systems, reopen closed paths without new evidence, or expand scope merely because a path is blocked.\n\nStop on missing authority, ambiguous state, destructive remediation without explicit approval, missing receipt, or unsatisfied production gate.\n"
}