@tuobadaidai

orchestration-reviewer-skill evaluates existing workflow plans against business goals across seven dimensions (alignment, topology, skill binding, scope safety, executability, efficiency, governance), outputting pass/reject verdicts with scorecard and remediation guidance.

skill-crafter guides creation of high-quality skills through four-layer structure and four-phase process (requirements understanding, authoring, validation, iteration). targets users building new skills rather than using existing ones.

resume-to-tags extracts atomic labels from resume text and builds a feishu multi-select table for talent matching. covers tag extraction, synonym expansion, table creation, bulk record insertion, and cleanup.

smart-search-fallback routes queries across seven tool tiers (serper, opencli google, crawl4ai, firecrawl, web-fetch, verticals) with automatic degradation on failure. covers english and chinese content via scenario-based tool selection.

diting-employee-development-expert covers talent inventory, executive assessment, succession planning, and career pathing via five structured scenarios with templated outputs, though lacks explicit failure modes and recovery paths for common implementation risks.

guanshi-industry-expert performs structured industry analysis across market sizing, porter's five forces, value chain mapping, and lifecycle assessment. designed as a cluster member dispatched by chief orchestrator, not for independent use or simple lookups.

guanshi-strategy-expert covers company-level strategic analysis including portfolio assessment, growth path design, execution decomposition, and second-curve identification. intended for dispatch via guanshi chief workflow, not standalone invocation.

agent-optimization-expert claims to diagnose and repair agent execution issues across openClaw and hermes platforms, but the procedures are vague checklists rather than actionable steps, and the claimed self-evolution mechanism lacks concrete implementation details.

diting-org-management-expert covers org design, team management, cross-functional collaboration, leadership development, and change management. scope is clear but procedures lack step-by-step rigor, relying instead on framework names and conceptual boxes.

diting-performance-expert covers kpi/okr design, target setting, performance improvement plans, and organizational health assessment. intended for enterprise performance management workflows within the diting chief orchestration layer, with explicit out-of-scope boundaries for compensation and labor law.

diting-training-expert covers training needs analysis, program design, kirkpatrick evaluation, and capability building across organizational talent development. scope clearly excludes compensation, compliance, performance management, and recruitment.

bazi-master applies classical Chinese metaphysics (滴天髓, 穷通宝鉴, 三命通会) to analyze birth charts. triggers on八字/命理/排盘queries, outputs narrative readings without exposing reasoning chains.

chief is an hr diagnostic framework claiming system-2 rigor via mckinsey seven-step method, but lacks executable procedure detail. the skill conflates architectural vision with operational reality: extensive state schemas and evaluator loops exist in prose form only, with no clear mapping to actual agent execution or error recovery.

深度洞察引擎 — 对任何分析结论执行穿透式深度审查，输出带有因果链、反事实场景和张力暴露的洞察报告。不与领域知识耦合，只负责分析方法论。Use when 需要深化分析结论、穿透表层判断、暴露战略建议的可执行性风险。不适用于领域数据查询、简单事实确认。

观势专家集群 — 竞争情报专家。Use when 被观势 Chief 调度执行竞争对手画像、竞争格局分析、对标分析、竞对战略意图推断、竞争脆弱性评估、博弈推演. 不适用于简单竞对信息查询或独立调用.

观势专家集群 — 财务战略专家。Use when 被观势 Chief 调度执行财务模型搭建、投资回报分析、企业估值、资本配置策略、风险量化、现金流压力测试、并购财务评估. 不适用于个人理财、日常报销或独立调用.

观势专家集群 — 组织诊断专家。Use when 被观势 Chief 调度执行组织架构分析、人才梯队评估、文化与战略匹配诊断、变革准备度评估、组织能力差距分析. 不适用于 HR 操作类问题（如请假/入职流程）或独立调用.

观势专家集群 — 市场洞察专家。Use when 被观势 Chief 调度执行用户研究、需求分析、市场细分、定价策略、市场进入策略、客户旅程分析、品牌定位. 不适用于市场调研数据查询或独立调用.

谛听薪酬专家。薪酬对标、调薪方案、薪酬公平性、薪酬结构设计。Use when 薪酬对标、调薪方案、薪酬公平性分析、薪酬结构设计、市场分位值分析、薪酬带宽设计、薪酬竞争力评估、宽带薪酬设计。不适用于绩效体系设计、劳动法合规、培训方案设计等非薪酬类 HR 问题。

谛听劳动法规专家。专注中国劳动法合规与用工风险管理。Use when 劳动法咨询、用工风险评估、员工辞退、劳动仲裁应对、竞业限制、试用期管理、加班合规、跨国用工、劳动合同解除、经济补偿计算。不适用于薪酬结构设计、绩效体系设计、培训方案设计等非法规类问题。

谛听AI应用专家。AI场景设计、工具选型、变革管理、数据隐私、AI政策、ROI评估。Use when AI应用落地、智能工具选型、AI变革管理、员工AI素养培训、AI安全合规、AI投资回报分析。不适用于纯技术开发、算法优化等非业务应用场景。

谛听行政专家。行政流程优化、办公环境管理、供应商管理、活动策划、员工服务、智能行政。Use when 行政流程优化、办公环境规划、供应商评估、年会团建策划、差旅政策设计、采购管理、员工服务提升、AI 行政优化、智能工单、自动化办公。不适用于人力资源政策设计、薪酬绩效、劳动法合规等非行政类问题。

将复杂业务目标编译为可静态校验的工作流执行计划（Plan IR）。 Use when user asks to 编排多步任务、规划工作流、生成DAG执行计划、 多Skill编排、任务拆解为DAG、生成结构化执行计划、workflow planning. 不适用于单轮问答/单工具操作/纯文本改写/轻任务。

基于简历、面试记录和JD，对候选人进行AI时代能力定级（L1-L4）。 核心能力：简历漏洞穿透审计、六维度评分卡、双乘数加权、测谎面试题生成、评分一致性校准、 面试认知复盘（v3.3 Pro 新增）。v3.3 从"简历打分器"升级为"认知行为分析系统"， 评估候选人如何思考、如何应对不确定性、如何与AI协同。 U...

薪酬市场调研分析技能（增强版）。支持多源数据收集、24字段标准化CSV模板、年薪自动计算、分位值计算（P10-P90）、17章节专业报告生成、多格式输出（Excel/Markdown/Word）。

薪酬市场调研分析技能（全球版）。支持国内+海外多源数据、28字段全球化模板、多币种自动换算（20+货币）、PPP购买力平价调整、分位值计算、17+章节专业报告、外派薪酬建议、多格式输出。

Agent 自愈引擎 — 自动诊断和修复 OpenClaw 执行问题（Cron 失败、工具报错、工作流中断、性能退化），自带持续自我进化能力。 Use when user asks to 系统诊断、修复 cron、优化配置、检查系统健康度、自愈修复、诊断失败原因、修复任务失败、优化师自我更新. 不适用于业务逻辑问...

工资审核 AI 技能。基于审核月份、区域范围、特殊情况，一次性自动生成完整的工资审核报告。 覆盖 10 项核心审核要素（人员范围/入离职/异常检查/考勤/绩效/奖金/社保/个税/人才房/计算逻辑），支持国内和海外员工审核。 三路输出：Markdown 审核报告 + 飞书文档 + HTML 可视化报告（含图表、审核...