Files
ai-town/engine-python/app/action_points.py
empty 5ae63d9df9 feat(engine): 实现行动点系统
- 扩展 AgentState 添加 action_points, max_action_points, last_action_tick 字段
- 新增 ActionFeedback 模型用于返回行动执行结果
- 创建 action_points.py 模块实现行动点消耗与恢复逻辑
- 行动消耗表: vote=1, trigger_skill=2, influence=2, comment/support/chaos=0
- 每 tick 恢复 1 点行动点(不超过 max)
- 行动点不足时拒绝执行并返回失败反馈
- 新增 7 个测试用例,全部 37 个测试通过

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-30 13:48:33 +08:00

90 lines
2.6 KiB
Python

"""行动点系统 - 管理用户行动点的消耗与恢复"""
from typing import List, Tuple, Dict
from .models import WorldState, Event, ActionFeedback
# 行动消耗表
ACTION_COST: Dict[str, int] = {
"vote": 1,
"trigger_skill": 2,
"influence": 2,
"comment": 0,
"support": 0,
"chaos": 0,
}
DEFAULT_COST = 0
def get_action_cost(event_type: str) -> int:
"""获取行动消耗的点数"""
return ACTION_COST.get(event_type, DEFAULT_COST)
def check_action_points(state: WorldState, user: str, cost: int) -> bool:
"""检查用户是否有足够的行动点"""
if user not in state.agents:
return True # 非 agent 用户不受限制
return state.agents[user].action_points >= cost
def consume_action_points(
state: WorldState, user: str, cost: int
) -> None:
"""消耗行动点"""
if user not in state.agents:
return
agent = state.agents[user]
agent.action_points = max(0, agent.action_points - cost)
agent.last_action_tick = state.tick
def regenerate_action_points(state: WorldState) -> None:
"""每 tick 恢复行动点"""
for agent_id, agent in state.agents.items():
if state.tick - agent.last_action_tick >= 1:
if agent.action_points < agent.max_action_points:
agent.action_points = min(
agent.max_action_points,
agent.action_points + 1
)
def process_event_with_ap(
state: WorldState, event: Event
) -> Tuple[bool, ActionFeedback]:
"""处理单个事件的行动点检查
返回: (是否允许执行, 反馈信息)
"""
user = event.user
cost = get_action_cost(event.type)
# 0 消耗的行动不需要检查
if cost == 0:
return True, ActionFeedback(
success=True,
reason="action applied",
remaining_ap=state.agents[user].action_points if user in state.agents else 0,
user=user
)
# 检查行动点
if not check_action_points(state, user, cost):
remaining = state.agents[user].action_points if user in state.agents else 0
return False, ActionFeedback(
success=False,
reason=f"insufficient action points (need {cost}, have {remaining})",
remaining_ap=remaining,
user=user
)
# 消耗行动点
consume_action_points(state, user, cost)
remaining = state.agents[user].action_points if user in state.agents else 0
return True, ActionFeedback(
success=True,
reason="action applied",
remaining_ap=remaining,
user=user
)