$ loading_
Prioritize churn reviews with transparent evidence and suggested next steps.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "churn-prediction" skill from askskill: 1. Download https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/solutions/customer-sentiment-churn/manual/skills/aibast_churn-prediction_02/SKILL.md 2. Save it as ~/.claude/skills/aibast_churn-prediction_02/SKILL.md 3. Reload skills and tell me it's ready
Using the churn-review prioritization rule, tell me which customer my team should review first today and list the evidence behind that priority. Separate observed evidence, heuristic output, and proposed next steps.
Returns the customer or record that should be reviewed first, with source-backed evidence, heuristic reasoning, and suggested next steps.
Analyze churn-review priority only for the specified fictional record CUST-8004 and do not substitute another record. Output the evidence, priority conclusion, and what authorized human review is required before action.
Provides a priority conclusion for that record, cites the evidence source, and states that human review is required first.
Produce a churn-review prioritization note that includes observed evidence, heuristic output, and next-step suggestions, and state that this is not legal, regulatory, insurance, lending, tax, investment, or financial advice, and that no external action occurred.
Gets a clearly structured review note with the required disclaimers.
When a retention or support team needs to decide which customer records to review first, this skill can rank them with a transparent heuristic and show the evidence behind the priority. It emphasizes human review before any follow-up action.
When the team wants to check one specific fictional record, it returns source-backed evidence for that exact record and avoids substituting another one. The output separates factual evidence, heuristic results, and suggested steps.
When an internal note is needed, it can include required statements such as that the result is not various forms of professional advice and that no payment, transaction, or external communication occurred. This is useful for documenting review conclusions cautiously.
The document describes a churn-review prioritization skill that uses a transparent heuristic to decide which records should be reviewed by humans first, rather than predicting an individual outcome. It outlines a fixed procedure: identify the exact record or scope, use the synthetic operating snapshot for source-backed evidence, separate observed evidence from heuristic output and next steps, include required disclaimers, and name the authorized human review needed before action. A locked example references a Retention Specialist and record CUST-8004.
Applies a transparent heuristic to prioritize human review without predicting an individual outcome.
Persona: Retention Specialist
Prompt: Who should my team review first today, and what evidence drove the priority?
Expected synthetic evidence: CUST-8004, prioritize review.
No. The document says it applies a transparent heuristic to prioritize human review rather than predicting an individual outcome.
It should identify the exact record or report scope, use the synthetic operating snapshot for source-backed evidence, and separate observed evidence, calculated or heuristic output, and next-step suggestions. It also needs to state that it is not various forms of professional advice and that no external action occurred.
It must name the authorized human review required before action. The document does not provide more detailed installation or runtime requirements.
Rank a fictional contract portfolio to prioritize legal review attention.
Explain synthetic cross-sell affinity rules and benchmark assumptions from fixed evidence.
Compare contract clauses against internal policy requirements and identify gaps.
Draft prioritized amendment positions and legal escalation points for contract renegotiations.
Scan synthetic customer records to find evidence-backed cross-sell product gaps.
Draft prioritized cross-sell recommendations and a reviewable engagement plan from synthetic data.
Prepare reviewable retention and service-recovery options for churn-related customer cases.
Aggregate customer feedback into theme insights and weekly action priorities.
Review pipeline health, surface risks, and plan weekly deal priorities.
Identify top fixable customer issues and draft response templates from support data.
Review product metrics, spot trends, and get actionable improvement recommendations.
Review contracts against playbooks with redlines, deviation flags, and impact analysis.