executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
combat_priority_scorerRank simplified AutoAscend combat actions with symbolic safety, damage, and line-of-fire checks.
Independent GitHub snapshot · Automated classification · Original documentation governs
Use this skill when a symbolic NetHack controller needs a deterministic combat-action ranking. It preserves AutoAscend’s paper mechanism: reduce combat to a smaller action set and score actions by safety and tactical value.
hero and monsters fields.selected_action.ranked_actions, each with score and reasons.Run python scripts/combat_scorer.py state.json --output combat.json or the tests. The tests verify low-HP healing, floating-eye avoidance, and line-of-fire penalties.
The scorer is a reusable symbolic approximation. It does not calculate full NetHack damage tables or ray-reflection probabilities.