$ loading_
Generate sentiment splits and recent excerpts from synthetic records.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "sentiment-analysis" skill from askskill: 1. Download https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/solutions/product-feedback-synthesizer/manual/skills/aibast_sentiment-analysis_03/SKILL.md 2. Save it as ~/.claude/skills/aibast_sentiment-analysis_03/SKILL.md 3. Reload skills and tell me it's ready
Using the uploaded synthetic records, give me the sentiment split and recent excerpts that support the result.
Return positive/negative split, supporting excerpts, and note that it is synthetic decision support.
For a product director, turn the sentiment analysis into a reviewable brief: lead with evidence, then explain uncertainty and the next review prerequisites.
Produce a reviewable brief that leads with evidence and is ready for review.
A product director needs a quick view of whether feedback is positive or negative, plus recent evidence excerpts. The result should answer directly and make clear the data is synthetic.
When someone needs a draft before a team review, this skill can lead with evidence and then state uncertainty and next review prerequisites. It does not make decisions or perform external actions.
The README says this skill works only with uploaded synthetic records and rules. It should lead with evidence that answers the product director’s question, then explain uncertainty, prerequisites, and the authorized review needed next. The result must be framed as synthetic decision support, with no external action taken, no system changes, and no transactions completed.
Use when a director of product asks for the fictional sentiment split and recent excerpts.
PositiveNegativefictional pilot dataDo not claim to have changed a system, contacted a person or supplier, made a decision, or completed a transaction. Stop at a reviewable brief or draft.
It generates a sentiment split and recent excerpts from uploaded synthetic records and rules for product decision support. The result must state that it is synthetic decision support and that no external action occurred.
No. The document says to use only the uploaded synthetic records and rules, and not to claim contacting anyone, changing systems, or completing transactions.
Lead with the specific evidence that answers the question, then explain uncertainty, prerequisites, and the next authorized review. The result should stop at a reviewable brief or draft.
Find each line’s bottleneck and recommend which station to address first.
Generates controlled implementation checklists for human review.
Get a synthetic health overview of a loyalty program without changing it.
Review SIU indicators in claims and return explainable evidence with human review guidance.
Analyze loyalty tier thresholds, perks, progress, and tradeoffs without changing status.
Identifies tax-loss-harvesting candidates and required review controls.
Summarize feedback volume, channels, categories, and commercial context.
Analyze meetings to reveal communication patterns and actionable improvement feedback.
Turn raw research feedback into structured insights and prioritized recommendations.
Generate actionable brand voice guidelines from documents, calls, or discovery reports.
Prepare reviewable retention and service-recovery options for churn-related customer cases.
Synthesize user research into themes, insights, and prioritized recommendations.