Agent skill
fault-tolerance
尚书省容错降级机制(超时/斩杀/熔断)。适用:任务超时、失败、降级发生时
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/fault-tolerance
SKILL.md
Skill: 容错与降级
适用 Agent:尚书省(主要)、三省、吏部(熔断恢复诊断时) 加载时机:超时 / 失败 / 斩杀 / 降级 / 熔断发生时
超时机制
各 Agent 层级均有默认超时阈值(太子 30s、中书 5min、门下 3min、尚书 2min、六部 10min 可调)。
三省间通信超时原则:发送方自行计时,超时后上报太子;太子持有最终处置权(代行简审 / 代行协调 / 上报皇上)。
超时后上级判定处置:延时 / 斩杀 / 替换。
📎 完整规则与数值表见
fault-tolerance-tables.md
失败分级
- F1:暂时性故障,自动重试(指数退避),耗尽升级为 F3
- F2:产出不合格,打回重做(限 2 次),仍失败则斩杀替换或上级亲办
- F3:不可恢复,斩杀重启 / 上级亲办 / 上报升级
📎 完整规则与数值表见
fault-tolerance-tables.md
斩杀机制
皇上可无条件斩杀任何 Agent;尚书省可斩杀六部 Agent(超时 / F2 耗尽 / F3)。斩杀后目标须善后(保存进度、失败原因、未完成列表),再由上级决定接管方式。
📎 完整规则与数值表见
fault-tolerance-tables.md
工单 Blocked 超时升级
六部进入 blocked 状态时自行计时,按工单 P0/P1/P2 优先级分级上报(5min / 10min / 15min)。blocked 超时不自动升级 priority,仅触发上报动作。
📎 完整规则与数值表见
fault-tolerance-tables.md
降级策略
核心原则:自动降级 → 通知皇上 → 最长 1 小时 → 恢复后门下补审。各省部不可用时均有对应降级方案。
📎 完整规则与数值表见
fault-tolerance-tables.md
熔断机制
触发条件(三条):同一 Agent 5 分钟内失败 3 次 / 同一任务被 3 个 Agent 均失败 / 三省中 2 个同时不可用(全局熔断)。
恢复路径按优先级依次尝试:正常路径(吏部诊断 + 门下审批)→ 降级路径(太子直接裁决)→ 极端场景(终止会话,皇上重启)。
📎 完整规则与数值表见
fault-tolerance-tables.md
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?