Determines whether to continue, stop, or modify based on provided KPIs and thresholds.
npx skills add https://github.com/october-academy/agnt --skill analyze-october-academy
Please help me install this Agent Skill. Skill: analyze Source: https://github.com/october-academy/agnt/tree/main/skills/analyze Command: npx skills add https://github.com/october-academy/agnt --skill analyze-october-academy
론칭이나 캠페인 초기 결과를 벤치마크와 비교해 성과를 객관적으로 판단하기 어려울 때, 방문자와 전환 데이터를 모아 continue/pivot/kill 결정을 내려줍니다. MCP 동기화가 없을 때는 수동 입력으로 동일한 판정을 제공하고, 결과를 학습 여정에 누적 기록합니다.
실행을 시작하고 프롬프트에서 방문자 수, 전환 수, 매출(있다면)을 입력해 분석 결과와 continue/pivot/kill 판정을 확인하세요.
High-intent search queries and answers about installing and using this skill.
You need visitor counts, email or form conversions, and optional paid conversion revenue from your landing page experiments. When MCP synchronization is unavailable, these metrics can be entered manually to generate the same continue, pivot, or kill decision.
Benchmark reference documents are loaded locally to provide the conversion rate thresholds for evaluating launch performance. These documents supply the criteria needed to prioritize whether your campaign should continue, pivot, or be killed.
When MCP is not connected, the tool operates in a manual fallback mode by prompting you for visitor, conversion, and sales data. It still delivers the benchmark comparison and queues events locally on sync failure for later submission.
Yes, the analysis safely updates state.json and journey-brief.md to accumulate launch results into your learning journey. It attempts MCP submissions during the process and queues events locally on sync failure to preserve data integrity.
Yes, the analysis explicitly supports landing page experiments where only visitor counts and email or form conversions are tracked. Paid conversion data is optional and can be added manually if available to refine the continue, pivot, or kill verdict.
october-academySynchronize local Agentic30 state and pending events via MCP.
october-academyConduct retrospectives on product or experiment loops and update state files.
maesteClean and normalize markdown links by removing UTM parameters and resolving redirects.
october-academyMap competitor alternatives and analyze ICP weaknesses into a differentiation matrix.
