Agent Skills › screenpipe/screenpipe › screenpipe-interview-synthesis

screenpipe-interview-synthesis

GitHub

分析单次发现型访谈，提取用户行为、需求、变通方案及未解问题。通过 API 获取记录，区分实际行为与主观陈述，结合数据验证结果，输出证据充分的总结及后续调研建议。

crates/screenpipe-core/assets/skills/screenpipe-interview-synthesis/SKILL.md screenpipe/screenpipe

Trigger Scenarios

需要分析单次用户访谈记录从访谈中提取具体行为和痛点生成基于证据的访谈总结报告

Install

npx skills add screenpipe/screenpipe --skill screenpipe-interview-synthesis -g -y

More Options

Non-standard path

npx skills add https://github.com/screenpipe/screenpipe/tree/main/crates/screenpipe-core/assets/skills/screenpipe-interview-synthesis -g -y

Use without installing

npx skills use screenpipe/screenpipe@screenpipe-interview-synthesis

指定 Agent (Claude Code)

npx skills add screenpipe/screenpipe --skill screenpipe-interview-synthesis -a claude-code -g -y

安装 repo 全部 skill

npx skills add screenpipe/screenpipe --all -g -y

预览 repo 内 skill

npx skills add screenpipe/screenpipe --list

SKILL.md

Frontmatter

{
    "name": "screenpipe-interview-synthesis",
    "description": "Analyze one discovery interview for observed behavior, needs, workarounds, and unanswered questions."
}

interview synthesis

Identify the interview and research question. Use screenpipe-api to retrieve its note and bounded transcript. Check speaker identity and recording coverage before attributing statements or quoting.

Capture concrete past behavior: the situation, intended outcome, steps taken, current workaround, constraints, and consequences. Separate what the participant did from what they say they might do, interviewer suggestions, and your interpretation. A positive reaction is not adoption or purchase intent.

When comparing the interview with usage data, name the outcome the participant wanted and check whether the measured action demonstrates it. Retrieval or a generated draft alone does not prove that outcome. State missing coverage and look for completion in the requested destination when access is available.

Return a concise evidence-backed account with source pointers, conflicting statements, and the next questions to investigate. One participant does not establish prevalence. Use role labels in shareable drafts and preserve identifying quotes only when appropriate to the requested audience. For themes across multiple interviews, use screenpipe-research-synthesis.