Skip to content

Instantly share code, notes, and snippets.

@johnsoncheg
Last active March 31, 2026 01:40
Show Gist options
  • Select an option

  • Save johnsoncheg/892d2f8e86eabae3d1c7bd6a01c3e1b2 to your computer and use it in GitHub Desktop.

Select an option

Save johnsoncheg/892d2f8e86eabae3d1c7bd6a01c3e1b2 to your computer and use it in GitHub Desktop.
续保车险skill

很简单,你可以把它理解成一个**“车险续保对比和压价提示词模板”**。 使用方式就是:以后你把销售聊天记录、报价单截图、福利信息发给装了这个 skill 的助手,它就会按这个套路自动帮你拆。

你可以怎么用

最常见的就是这几种问法:

用法 1:让它直接分析报价

你可以说:

我这边有 3 个销售的续保报价,你帮我按整单口径对比一下,先确认是否包含交强险,再算最终现金实付和综合价值,并告诉我下一句该怎么回复压价。

───

用法 2:发单个销售,让它帮你拆

比如:

这是平安一个销售给我的报价单和聊天记录。 你帮我判断: 1)有没有附加包可以删 2)这是不是整单价格 3)最终实际要付多少 4)我下一句该怎么回,争取更低价和更多权益

───

用法 3:让它直接生成回复话术

比如:

这个销售说今天下单返 900,再送 500 盒马和 100 话费,但我怀疑他报价里有尊享包。 你按车险续保谈判模式帮我写一段回复,目标是确认是否含交强险、去掉附加包、保留优惠。

───

一个完整示例

你这样发:

我的是 2025 款特斯拉焕新 Y,首年续保。 A 销售是平安,报价总价 6164.81,交强险 855,商业险 4653.81,尊享保障 656。 她说返佣 900,再给 500 盒马 + 100 话费。 你帮我看看这单值不值,哪些该砍,我该怎么回她。

理想输出会像这样:

• 先判断:这是整单口径,含交强险 • 拆险种: • 核心主险正常 • 656 尊享保障属于优先审查项 • 算账: • 现金实付约 5264.81 • 另有 600 券包 • 结论: • 不是最低现金方案 • 但综合价值还行 • 然后给你一句可复制回复:

我主要按最终总实付比较,麻烦你帮我把尊享保障去掉后再出一版,同时确认 900 返佣和券包是否还能保留,我这边按整单净价一起比较。

───

最适合这个 skill 的场景

它尤其适合处理这几类情况:

1)多家保险公司、多名销售同时报价

比如:

• 平安 2 个 • 太平洋 3 个 • 人保 1 个

2)销售喜欢把福利说得很花

比如:

• 超充 • 洗车 • 喷漆 • 代驾 • 机场停车 • 红包 • 返佣 • 话费券

这时候 skill 会帮你统一折算,不容易被带偏。

3)你想让助手直接帮你谈判

也就是:

• 帮你判断谁更强 • 帮你生成下一句 • 帮你识别销售是不是在偷换口径

───

最关键的一个使用习惯

你以后每次用这个 skill,最好都带一句:

先帮我确认是否包含交强险,再比较最终总实付。

因为这就是这次实战里最重要的教训。

最短使用示例

如果你想一句话启动它,直接说:

帮我按车险续保谈判模式分析这份报价:先确认是否含交强险,再拆附加项、算现金实付,并给我一句可以直接发给销售的话。


标准提问模板

请按“车险续保谈判/比价”模式帮我分析下面这份报价,目标是: 1)争取最低总实付 2)剔除无用附加险/打包项 3)争取最大福利(返现、返佣、券包、实用权益)

分析时请务必按这个顺序来:

  • 先确认这份报价是否包含交强险、车船税,还是只是商业险
  • 再拆主险、附加险、打包项
  • 再计算“现金实付”和“综合价值”
  • 最后给我一句可以直接发给销售的话,帮我继续压价或确认方案

这是我的背景:

  • 车型:
  • 续保年份:
  • 其他前提:

这是销售信息:

  • 保险公司:
  • 销售姓名:

这是聊天记录/报价信息: 【把聊天记录、报价单内容、截图OCR内容贴这里】

如果有图片,请你结合图片一起分析。 请按下面格式输出: 1)快速结论 2)这份报价是否为整单口径(是否含交强险) 3)核心险种拆解 4)附加项/打包项是否建议删除 5)现金实付 6)综合价值 7)风险点/话术陷阱 8)我下一句该怎么回复

───

你也可以用这个“多人对比版”

如果你以后一次发好几家销售,用这个更合适:

请按“车险续保谈判/比价”模式帮我对比下面几家销售的报价。

目标: 1)比较谁的最终总实付最低 2)识别谁塞了无用附加险/打包项 3)比较返现、返佣、券包和实用权益 4)给我每个销售下一句该怎么回 5)最终给出排名

注意:

  • 一定先确认每家的报价口径是否一致
  • 先看是否包含交强险、车船税
  • 如果有的只是商业险报价,请明确标注“暂不可直接横比”
  • 不要把不确定能用掉的权益按满值计算

我的背景:

  • 车型:
  • 续保年份:
  • 特别在意的权益:(例如机场停车、超充、充电券)
  • 不需要的权益:(例如洗车、代驾)

销售1:

  • 保险公司:
  • 销售姓名:
  • 聊天记录/报价:

销售2:

  • 保险公司:
  • 销售姓名:
  • 聊天记录/报价:

销售3:

  • 保险公司:
  • 销售姓名:
  • 聊天记录/报价:

请按以下格式输出: 1)总排名 2)每家是否整单口径 3)每家现金实付 4)每家综合价值 5)每家的主要风险点 6)每家下一句怎么回 7)推荐优先谈哪家

───

最短一键版

如果你懒得写很多,就用这句:

帮我按车险续保谈判模式分析这份报价:先确认是否含交强险,再拆附加项、算现金实付和综合价值,最后给我一句可以直接发给销售的回复。

───

我建议你再固定加一句

每次都加上这句最稳:

注意:如果报价没明确写交强险,请不要直接和其他整单报价横向比较。

这句非常值钱,因为它能防止重蹈这次 C 的坑。

name insurance-renewal-negotiation
description Compare and negotiate auto insurance renewal quotes across multiple sales reps or insurers, especially when the user shares quote sheets, screenshots, rebate offers, coupon bundles, and service perks. Use when the goal is to minimize final out-of-pocket cost, remove low-value add-ons, normalize quote scope (especially whether compulsory traffic insurance/交强险 is included), rank options by net value, and draft reply messages to push for lower price or better benefits.

Insurance Renewal Negotiation

Overview

Use this skill to normalize competing insurance quotes, identify hidden add-ons, calculate real net cost, and produce concise reply messages that push sales reps toward lower cash cost or more practical benefits.

Core rule: compare on the same scope first

Never compare quotes until the scope is normalized.

Check these items first:

  • whether the quote includes 交强险
  • whether it includes 车船税
  • whether the price is commercial-only or the full policy total
  • whether rebates are direct deduction, post-payment cash transfer, red packet, or coupon value
  • whether perks are usable/app-issued/store-registered or just verbal promises

If scope is inconsistent, say so explicitly before ranking.

Workflow

1. Build a quote card for each sales rep

For each rep, extract and track:

  • insurer
  • rep name
  • vehicle / renewal context
  • total quoted premium
  • compulsory insurance (交强险)
  • commercial premium subtotal
  • main coverages:
  • vehicle damage / 车损险
  • third-party liability / 三者险
  • medical outside social insurance for third party / 医保外医疗责任险(三者)
  • EV external power grid loss / 外部电网故障损失险
  • add-on bundles:
  • 驾乘险
  • 车上人员险
  • 尊享保障 / 乐驾人生 / 随行e / similar bundles
  • rebate / cashback / red packet / coupon bundle
  • perks and how they are fulfilled
  • key conditions such as “today only”, “Tesla-only”, “must include ride coverage”, “must pay first then rebate”

2. Mark removable items

Treat these as default review items, not automatic must-buys:

  • bundled personal accident / medical packages
  • ride/passenger accident packages
  • family property / health / medicine / specialty drug bundles
  • broad “premium owner package” products
  • duplicated non-core lifestyle coverage

Prefer keeping the core structure lean:

  • 交强险
  • 车损险
  • 三者险
  • 医保外医疗责任险(三者)
  • EV-specific low-cost practical items when relevant

3. Calculate two numbers

For each quote, compute:

  1. Cash out-of-pocket
  • the money the user actually pays after confirmed direct deductions or confirmed cash-back
  1. Net value view
  • cash out-of-pocket minus only the perks the user is very likely to use

Do not value coupons/services at face value by default. Use conservative assumptions.

Suggested treatment:

  • cash rebate / direct deduction: count at 100%
  • general phone credit / highly usable grocery coupon: count near 100% if user will use it
  • app-issued charging / parking / ride perks: count only if user says they matter
  • wash / detailing / spray-paint / store-limited services: count low or zero unless the user wants them
  • claims convenience / rescue / fast-pay: operational value, usually do not overprice

4. Rank with clear caveats

Use this ranking order:

  1. lowest confirmed full-policy cash out-of-pocket
  2. cleanest coverage structure
  3. best practical perks
  4. easiest/clearest fulfillment path

If a quote is missing 交强险 or another major component, mark it as:

  • not directly comparable yet
  • commercial-only for now

5. Draft the next reply

After analysis, draft a short reply optimized for one of these goals:

  • ask for a normalized quote
  • remove add-ons
  • confirm rebate mechanics
  • convert weak perks into practical ones
  • use another quote as pressure without oversharing full details
  • move toward final confirmation

Negotiation heuristics

Ask these questions early

  • Does this include 交强险?
  • Is this the full-policy total or only commercial insurance?
  • What is the final cash amount I actually pay?
  • Is the rebate direct deduction or post-payment transfer?
  • When does the rebate arrive?
  • Which perks are guaranteed, and how are they issued?
  • If I remove the add-on package, how much does the premium drop?

Good pressure lines

  • “我主要按最终总实付比较。”
  • “你先按整单口径给我,我只看同条件下的净价。”
  • “附加包先去掉,我先看核心责任最低价。”
  • “这些权益如果我用不上,能不能换成更实用的权益或者折现一点?”
  • “如果你这边今天能给到最好方案,我可以优先考虑。”

Avoid these mistakes

  • do not rank quotes before confirming whether 交强险 is included
  • do not treat coupons as cash unless the user will truly use them
  • do not assume a verbal “actual payment” number is final without asking how it is calculated
  • do not overvalue bundled perks tied to specific stores unless the user wants them
  • do not let a rep steer the discussion away from full-policy total

Output pattern

When replying to the user, prefer this structure:

Quick conclusion

  • who currently leads and why

Quote breakdown

  • total
  • cash rebate
  • estimated cash out-of-pocket
  • notable add-ons
  • practical perks
  • fulfillment risk

Recommended reply

  • 1 short message the user can copy to the rep

Watch-outs

  • missing scope items
  • inconsistent conditions
  • “today only” pressure
  • post-payment rebate risk

Special lesson from real use

If a quote appears dramatically cheaper, first verify whether it is missing 交强险. A commercial-only quote can look like the lowest cash payment while not actually being the cheapest full-policy option.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment