Extract, transform, analyze and visualize data from documents and structured sources.
Use when extracting text from scanned PDFs, photographed pages, or images that have no embedded text layer. Covers OCR backends, language packs, force-OCR, and performance tuning.
Use when extracting from many files at once with shared config, bounded parallelism, per-file overrides, and error recovery. Covers the `batch` command, `--file-configs`, `--max-concurrent`, and output layout.
Use when extracting keywords (YAKE/RAKE) from documents — and, secondarily, when detecting document language or generating embeddings for RAG and search. Covers the keyword config (and its feature gating), `--detect-language`, and the standalone `embed` command with real flags.
Aggregate and rank signals from multiple edge-finding skills (edge-candidate-agent, theme-detector, sector-analyst, institutional-flow-tracker) into a prioritized conviction dashboard with weighted scoring, deduplication, and contradiction detection.
This skill should be used when analyzing market breadth charts, specifically the S&P 500 Breadth Index (200-Day MA based) and the US Stock Market Uptrend Stock Ratio charts. Use this skill when the user provides breadth chart images for analysis, requests market breadth assessment, positioning strategy recommendations, or wants to understand medium-term strategic and short-term tactical market outlook based on breadth indicators. Also works WITHOUT chart images by fetching CSV data directly from public sources. All analysis and output are conducted in English.
上传一份年度报告 PDF,解读经营表现、利润质量、现金流与资产负债,并关联原文依据。
基于已连接的业务数据源构建有数据证据支撑的经营仪表板,监控核心指标、跟踪变化并定位变动因素,输出可恢复的看板草稿、指标口径与数据证据。当用户要求做仪表板、看板、大盘、监控面板、经营驾驶舱或持续跟踪指标时使用。
数据库健康巡检技能。用 db_metrics 建立连接数、QPS/TPS、慢查询率、缓冲池命中率、磁盘空间与 IO、复制延迟、锁等待等多维基线,对比阈值与历史趋势,输出带分级与整改建议的巡检报告。触发词包括"巡检""健康检查""体检""健康度""数据库状态""水位""容量""连接数打满""复制延迟""主从延迟""磁盘快满""缓冲池命中率""例行检查""inspection""health check"。全程只读,不执行任何变更。
Build physical ER diagrams from declared keys and foreign keys. Use for schema relationships or ER documentation. Do not present inferred joins as lineage.
数据库画像与治理技能。生成目标库的结构与容量画像(表体积/行数分布、膨胀与碎片、无主键表、大字段与 TEXT/BLOB 滥用、冷热数据分布、索引冗余、命名与注释规范),并给出分层治理建议与归档/分区/拆分路线图。触发词包括"profile""画像""数据库画像""治理""数据治理""表膨胀""碎片""碎片率""无主键""大字段""冷热分层""归档""历史数据""容量规划""分库分表""瘦身""优化存储""库太大"。只出画像与提案,不自动执行任何变更。
上传两到五份年度报告 PDF,比较同年多家公司或同公司跨年的表现,解释可比口径和数据缺口。
Run a structured health patrol over a MySQL instance — resources, connections, locks, replication, space, schema risks and config drift — producing a graded report with an actionable issue list and owners.
Review MySQL configuration and resource allocation against the actual workload — memory budget, redo/undo, flush policy, connections, IO settings and 8.0-specific switches — with expected effect, risk and rollback for each change.
智能分析场景的数据集圈定技能。基于语义层表白名单与血缘 join 路径把分析范围固化为数据集卡(dataset_card),禁止从全量库里猜表选表。当用户要求"圈数据集/确定分析范围/看这次分析用哪些表/绑定表范围与口径/切换数据集/数据集太大或口径不清",或在开始归因、趋势、对比、报告类分析前需要先确定表范围、join 路径、默认过滤与指标版本时使用。触发词:圈数据集、数据集、表范围、白名单、join、血缘、口径绑定、分析范围、dataset。
Handle a live MySQL incident — classify the symptom, walk the matching decision tree to find root cause, propose safe mitigation first, then produce a timeline-based postmortem with action items.
Build a physical data dictionary from authorized metadata. Use for field documentation or dictionary CSV export. Do not invent business definitions or scan business rows.
智能分析场景的归因分析技能。对指标异动(下滑/上涨/波动)做多维度贡献度拆解、环比同比对比、量价分解、结构效应与规模效应分离,并用排除法定位异动起点与影响范围,输出主因/次因/待验证假设三段式归因结论卡。当用户要求"归因/为什么下滑/为什么上涨/谁贡献了/异动定位/根因分析/贡献度拆解/环比同比分析/量价分解/复盘",或需要判断某个维度是否是主要原因、需要区分数据口径问题与真实业务变化时使用。触发词:归因、贡献度、拆解、为什么下滑、为什么上涨、异动、根因、环比、同比、量价分解、结构效应、复盘、attribute。
Read-only safe execution and cost guardrails. Use after a SQL passes sql.validate and needs to run on a read replica with cost estimation, approval routing (confirm on high-cost scans), and enforcement of timeout / row-limit / row-level-security / soft-delete filters; also handles timeout, over-row-limit, and approval-rejected exceptions.
Internal MySQL operations knowledge base — metric thresholds, symptom-to-root-cause decision trees, read-only diagnostic command library, and safety red lines. Referenced by other skills, not shown in the user menu.
索引优化建议技能。基于 index_info 的现有索引与基数、slow_log 的查询模式聚类、explain 的计划判读,生成组合索引候选、识别重复与未使用索引、评估写入放大与体积代价,并输出 ADD INDEX / DROP INDEX 提案(含在线 DDL 方案与回滚)。触发词包括"索引""加索引""建索引""索引优化""索引建议""组合索引""覆盖索引""冗余索引""重复索引""未使用索引""索引失效""该不该加索引""index advisor""ADD INDEX"。只出提案,绝不自动执行 DDL。