Build and Automate

Build a skill that runs every morning at 7:30am.
It should:
1. Read the latest bank-statement.csv and ledger-entries.csv from the workspace
2. Match entries by Reference_Number and Date
3. Separate into matched, variances, unmatched-bank, and unmatched-ledger
4. Produce a reconciliation report (.docx) with four sections, summary stats, and variance highlights
5. If any single variance exceeds 1,000 or total unmatched entries exceed 10, flag as CRITICAL and prepend a warning banner
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 7am.
It should:
1. Read all cases updated in the past 7 days from Jira
2. Group by scheme (Visa, Mastercard, AMEX)
3. Calculate approaching deadlines (< 7 days warning, < 3 days critical)
4. Produce a Word document (.docx) summary with deadline warnings
5. If any case is < 3 days from deadline, flag as CRITICAL in the summary
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 7am.
It should:
1. Read the daily screening alert export from the workspace (compliance-alerts.xlsx)
2. Categorise each alert as FALSE POSITIVE, NEEDS REVIEW, or ESCALATE based on similarity score and match type
3. Flag cross-alert patterns (same merchant appearing in multiple alerts)
4. Produce a triage summary (.docx) grouped by disposition with counts, escalation cases detailed, and pattern analysis
5. Any ESCALATE case must include the note: DRAFT — analyst must verify all match details before final disposition
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 8am.
It should:
1. Read country-metrics.csv from the workspace covering the past 7 days
2. Calculate week-over-week changes for sales, ticket volume, onboarding, and activation per country
3. Rank countries by performance and flag any metric that dropped more than 15% week-over-week
4. Produce a one-page digest (.docx) with tables, trend arrows, and a top wins / top concerns section per country
5. If any country shows three or more declining metrics simultaneously, flag as NEEDS ATTENTION at the top of the report
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 7:30am.
It should:
1. Read all Zendesk tickets received in the past 24 hours from support-tickets.xlsx
2. Categorise each as Onboarding Status, Terminal Issue, Settlement Query, Chargeback, Complaint, or Other
3. Assign priority (URGENT if merchant cannot process payments, HIGH if revenue impacted, NORMAL otherwise) and route to CR, Tier 2, Chargebacks, Retentions, or CDD
4. Produce a triage report (.docx) sorted by priority with suggested first response for each ticket
5. Any ticket categorised as Complaint must be flagged separately with a reminder to log a formal complaint record
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 6am.
It should:
1. Read the latest data-quality-checks.csv from the workspace (null rates, row counts, freshness per table)
2. Compare each metric against the 30-day rolling average and flag deviations greater than 2 standard deviations
3. Classify anomalies as STALE (freshness), VOLUME (row count shift), or INTEGRITY (null rate spike)
4. Produce an anomaly report (.docx) with a summary dashboard, per-table detail, and suggested investigation steps
5. If any anomaly affects a table feeding the commission pipeline or SAP integration, flag as FINANCIAL IMPACT and escalate to the Data team lead
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 8am.
It should:
1. Check for new settlement files in the workspace
2. Run a quick reconciliation against the internal ledger
3. Flag any discrepancies > 500
4. Produce a brief summary (.docx) — green if clean, yellow if minor issues, red if critical
5. Save to workspace with today's date in filename
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 7am.
It should:
1. Read merchant-chargeback-ratios.csv from the workspace covering the trailing 3 months
2. Calculate current chargeback ratio per merchant and compare against Visa VAMP and Mastercard MATCH thresholds
3. Flag merchants within 20% of either threshold and classify trend as increasing, stable, or decreasing
4. Produce a threshold monitoring report (.docx) with a summary table, flagged merchants detailed section, and 3-month trend charts
5. Any merchant that has crossed or is within 10% of a scheme threshold must be marked CRITICAL — senior fraud analyst review required before any action
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 7:30am.
It should:
1. Read hubspot-pipeline-export.xlsx from the workspace
2. Identify stale deals (not updated in 7+ days), stage bottlenecks, and top 10 deals closest to close
3. Flag at-risk deals where CDD is blocking or a competitor is mentioned in notes
4. Produce a pipeline health report (.docx) with three sections: stale deals with suggested actions, bottleneck analysis, and top 10 priority list
5. Deals stale for more than 14 days must be flagged as REQUIRES IMMEDIATE UPDATE — rep must add a note or close
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 8am.
It should:
1. Read software-licences.csv (tool, licence count, assigned users, last login date) from the workspace
2. Identify unused licences (no login in 30+ days) and orphaned accounts (user left company per HiBob export)
3. Calculate potential cost savings from reclaiming unused seats
4. Produce a licence audit report (.docx) with per-tool summary, unused seat detail, orphaned accounts, and estimated monthly savings
5. Any tool where unused licences exceed 20% of total must be flagged for manager review before reclamation
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 9am.
It should:
1. Read the JSM ticket export for Legal Help Desk requests from the past 7 days (legal-tickets.csv)
2. Categorise by type: NDA, contract review, GDPR request, regulatory submission, or other
3. Calculate ageing for each open request and flag any approaching SLA breach
4. Produce a legal workload report (.docx) with open items by category, ageing distribution, SLA risk items, and completed items summary
5. Any GDPR subject access request older than 20 days must be flagged as SLA CRITICAL — escalate to Data Privacy team immediately
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 7am.
It should:
1. Read delivery-queue.csv from the workspace covering all open terminal orders
2. Classify each order as On Track, Arriving Today, or Overdue based on expected delivery date
3. Count orders by type (new, swap, collection) and summarise returns pending QA
4. Produce a delivery status report (.docx) with today's schedule, overdue orders with courier details, and QA backlog
5. Any order overdue by more than 3 business days must be flagged as ESCALATE — contact courier and notify the merchant
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 8am.
It should:
1. Read campaign-results.csv from the workspace covering all active campaigns from the past 7 days
2. Calculate week-over-week change in open rate, click rate, conversion, and cost per acquisition per campaign
3. Flag any campaign with declining performance for two consecutive weeks
4. Produce a campaign performance digest (.docx) with per-campaign summary table, channel comparison, and top/bottom performers
5. If any campaign shows a conversion drop greater than 30% week-over-week, flag as REVIEW URGENTLY before next spend is committed
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 8am.
It should:
1. Read onboarding-pipeline.csv from the workspace listing all merchants in the onboarding journey
2. Identify the current stage for each merchant (Registration, KYC, CDD, Terminal, Activation) and calculate days in stage
3. Flag merchants as AT RISK if stuck in any stage for 14+ days, or BLOCKED if CDD pending for 21+ days
4. Produce a pipeline status report (.docx) with full pipeline table, bottleneck summary, and conversion funnel metrics
5. Any merchant in BLOCKED status for 28+ days must trigger a note: ESCALATE to Onboarding Lead — potential merchant loss
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 7:30am.
It should:
1. Read workato-failures.csv from the workspace listing all workflow automation bot recipe failures in the past 24 hours
2. Group failures by recipe name and error type, and count occurrences
3. Identify repeat offenders (recipes that failed 3+ times) and newly failing recipes (first failure in 7 days)
4. Produce a failure summary (.docx) with total failures, top failing recipes, error patterns, and suggested reprocessing steps
5. If any recipe failure affects merchant settlements or financial corrections, flag as FINANCIAL IMPACT — must be manually reprocessed before end of day
Use /skill-creator to build the skill definition.

Build a skill that runs on the 1st of each month at 8am.
It should:
1. Read commission-data.csv and partner-activity.csv from the workspace for the previous month
2. Calculate total commission per partner with breakdown by merchant, applying clawback rules for cancelled merchants
3. Flag anomalies: commissions changing more than 50% month-over-month, duplicate entries, or partners earning across multiple countries
4. Produce a commission report (.docx) with per-partner summary, detailed breakdown, anomaly section, and comparison vs prior month
5. Any flagged anomaly must include the note: REQUIRES SIGN-OFF — commission governance process mandates documentation before payment to SAP
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 8am.
It should:
1. Read upcoming-joiners.csv from the workspace listing all new joiners starting in the next 14 days
2. Check readiness for each: HiBob account created, IT ticket raised, equipment ordered, onboarding schedule drafted
3. Flag any joiner starting within 5 days who has incomplete items as AT RISK
4. Produce a new joiner readiness report (.docx) with per-joiner checklist, completion status, and department breakdown
5. If any joiner is starting in 2 or fewer business days with missing system access, flag as URGENT — escalate to IT Helpdesk immediately
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 8am.
It should:
1. Read activation-data.csv from the workspace for all merchants onboarded in the past 30 days
2. Identify merchants with no first transaction and classify the likely blocker (KYC incomplete, terminal not delivered, technical issue, or unknown)
3. Flag merchants 14+ days post-onboarding with no transaction as AT RISK
4. Produce a non-transacting merchant report (.docx) with full status table, blocker breakdown, and suggested outreach actions per merchant
5. Before sending any outreach, check with Sales to avoid duplicate messaging — include a coordination note on every merchant flagged for contact
Use /skill-creator to build the skill definition.

Build a skill that runs every Friday at 4pm.
It should:
1. Read jira-sprint-export.xlsx from the workspace for the current sprint
2. Calculate completed vs carried-over story points, bugs found vs fixed, and velocity trend over the last 4 sprints
3. Extract blockers and open dependencies from ticket comments and labels
4. Produce a sprint summary (.docx) with completion stats, carried-over items with reasons, blocker list, and next sprint risks
5. Any blocker that has been unresolved for more than 5 business days must be flagged as ESCALATE — PM to raise in next leadership sync
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 8am.
It should:
1. Read tier2-tickets-month.csv from the workspace covering all Tier 2 escalations from the past 7 days
2. Group tickets by issue category and count occurrences, comparing against the previous week
3. Identify the top 5 recurring themes with root cause hypothesis and affected merchant segment
4. Produce a recurring issue report (.docx) with pattern table, trend comparison, and product fix recommendations for each theme
5. Any issue category that increased more than 50% week-over-week must be flagged as TRENDING — notify Product team for prioritisation
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 7:30am.
It should:
1. Read merchant-activity.csv from the workspace covering transaction volumes, support tickets, and fee enquiries for the past 30 days
2. Flag merchants showing early warning signals: volume decline greater than 20% month-over-month, recent complaints, fee enquiries, or terminal issues
3. Rank flagged merchants by estimated churn probability based on signal count and severity
4. Produce a churn early warning report (.docx) with ranked at-risk merchants, signal detail per merchant, and suggested intervention (call, reprice, or escalate)
5. Any merchant with monthly volume above 50K showing 3+ warning signals must be tagged as HIGH VALUE AT RISK — Retentions lead must review before outreach
Use /skill-creator to build the skill definition.

Build a skill that runs every Monday at 7am.
It should:
1. Read merchant-risk-data.csv from the workspace with chargeback ratios, refund rates, and transaction patterns per merchant
2. Calculate distance from Visa VAMP and Mastercard MATCH thresholds for each merchant
3. Classify 3-month trend as increasing, stable, or decreasing and flag merchants within 20% of any threshold
4. Produce a threshold monitoring report (.docx) with summary dashboard, flagged merchants with full detail, and trend analysis
5. Any merchant that has breached or is within 10% of a scheme threshold must be marked as CRITICAL — senior risk analyst must review before any restrict or block action
Use /skill-creator to build the skill definition.

Build a skill that runs on the 5th of each month at 8am.
It should:
1. Read scheme-fees.csv (billed by schemes) and internal-estimates.csv (our calculations) from the workspace
2. Match by Scheme, Month, and Fee_Type, then calculate variance in amount and percentage
3. Flag any variance exceeding 5% or 1,000 and group discrepancies by scheme
4. Produce a fee reconciliation report (.docx) with matched summary, flagged discrepancies with detail, and per-scheme totals
5. Any discrepancy affecting interchange pass-through fees must be flagged as FINANCIAL REVIEW REQUIRED — Finance team must confirm before scheme is contacted
Use /skill-creator to build the skill definition.

Build a skill that runs every morning at 7am.
It should:
1. Read safeguarding-data.csv and fx-positions.csv from the workspace
2. Verify the 80/20 safeguarding split between partner bank A and partner bank B and flag any drift beyond 2 percentage points
3. Check each currency position against volatility thresholds and flag currencies approaching limits
4. Produce a morning treasury brief (.docx) with safeguarding summary, split check, FX exposure table, and DCC status per currency
5. If any currency has breached its volatility threshold or safeguarding split has drifted beyond 5 points, flag as IMMEDIATE ACTION REQUIRED — Treasury lead must review before market open
Use /skill-creator to build the skill definition.