AI Skills Directory

Batch convert timestamps between time zones. One timestamp per line; outputs a result table in txt/csv/json with ok/failed counts.

Extract document for office scenarios and output a structured, reviewable result with actionable recommendations. Prompt-class skill; results require human review. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Deterministic privacy-policy compliance check (PIPL/GDPR required sections). No LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Internal-control / SOX control-testing workpaper builder: reads a controls list, sizes samples by frequency and risk, and produces a control matrix, sampling plan, test worksheet template, deficiency-grading rubric, plus next-best-actions (actions.json) for high-risk controls.

Extract calendar for office scenarios and output a structured, reviewable result with actionable recommendations. Prompt-class skill; results require human review. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Media task skill: Music Lens (sub-variant mode x top_n; LLM-orchestrated). (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

One-command composite repo audit: file scope + code quality + dependency graph + dependency hygiene + data models + project context -> unified repo health report. No LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Domain LLM copilot for test review: pick an object (api/review) x operation and generate a structured artifact via runtime LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

28-mode analysis for Excel/CSV/SQLite: schema inspection, SQL query, statistical summaries, fraud detection (Benford, counterparty, network topology), large-amount test, Journal Entry Testing (JET), business-finance reconciliation (AR/AP aging, inventory turnover), and BI export.

Deterministic RFM customer segmentation. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Boss-ready weekly/monthly work summary: turns structured task records into a decision-oriented report -- overall progress, key items (goal/progress/metrics), risks needing coordination, next-period plan and decisions to escalate -- favouring data over filler and never hiding risk, with decision/follow-up next-best-actions (actions.json). For Chinese workplace (dev/sales/ops/PM/personal). (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Generates standardized work instructions (SOPs) from process parameters, equipment specs, and quality requirements for manufacturing operations.

Domain LLM copilot for marketing: pick an object (ad/brand/campaign/competitor/funnel/influencer/keyword/landing/market/newsletter/persona/positioning/product/promotion/review-text/sentiment/seo/social/tagline) x operation and generate a structured artifact via runtime LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Analyze survey response for generic scenarios and output a structured, reviewable result with actionable recommendations. Prompt-class skill; results require human review. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Deterministic cross-sell via product co-occurrence. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Deterministic exit-interview attribution. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Multi-engine web search: open-source, keyless DuckDuckGo (DDG Lite + duckduckgo_search library); merges and de-dups via reciprocal-rank fusion (RRF) with optional content extraction. Use it for topic research, fact-finding and competitor scans.

Domain LLM copilot for marketing: pick an object (ad/brand/campaign/competitor/funnel/influencer/keyword/landing/market/newsletter/persona/positioning/product/promotion/review-text/sentiment/seo/social/tagline) x operation and generate a structured artifact via runtime LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Analyze engagement survey: dimension scores, eNPS, drivers, by-department. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Merge multiple content files into one, or split a long document into parts by section or size. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Analyze sla for ops scenarios and output a structured, reviewable result with actionable recommendations. Prompt-class skill; results require human review. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Citation format proofing: normalize references to GB/APA style. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Domain LLM copilot for marketing: pick an object (ad/brand/campaign/competitor/funnel/influencer/keyword/landing/market/newsletter/persona/positioning/product/promotion/review-text/sentiment/seo/social/tagline) x operation and generate a structured artifact via runtime LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Data-driven weekly investment briefing over your watchlist or given symbols, exported as Markdown. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Orchestrate csv for office scenarios and output a structured, reviewable result with actionable recommendations. Prompt-class skill; results require human review. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Builds the three primary financial statements from a trial balance: income statement and balance sheet (fully computed, China-GAAP account-code mapping) plus an indirect-method cash-flow skeleton, with assets = liabilities + equity tie-out and review next-best-actions (actions.json).

Deep Q&A over an academic paper: scholarly snapshot, methodology x-ray, critical appraisal, plain-language explainer, grounded question answering. Use when you need deep q&a over an academic paper. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Reflow/convert SRT-VTT subtitles: line length, 2-line cap, format swap. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Domain LLM copilot for finance: pick an object (ar-aging/cashflow/consulting/cost-allocation/cost-behavior/depreciation/finance-fill/fx-exposure/goodwill/intercompany/investment/lc-document/multi-currency/petty-cash/related-party/revenue-recog/treasury/valuation-report/valuation-sens) x operation and generate a structured artifact via runtime LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Upload any document (PDF, scan, Word), name the fields you want, and it reads the text and extracts them into a structured table / JSON. Use it to parse contracts, pull fields from resumes, or extract part numbers and prices from quotes.