yanliudesign/offer-toolkit-skill / salary-negotiation-skill
salary-negotiation-skill
Verified| File | Kind | Size |
|---|---|---|
| .gitignore | other | 46 B |
| LICENSE | other | 1.0 KB |
| README.md | markdown | 9.0 KB |
| README.zh.md | markdown | 8.8 KB |
| SKILL.md | skill md | 12.1 KB |
| deal-bank/_deal-template.md | markdown | 2.6 KB |
| deal-bank/_index.md | markdown | 1013 B |
| examples/negotiation-report-template.html | other | 64.3 KB |
| frameworks/comp-benchmarks.md | markdown | 5.8 KB |
| frameworks/leverage-heuristics.md | markdown | 5.7 KB |
| frameworks/negotiation-report.md | markdown | 14.9 KB |
| frameworks/negotiation-tactics.md | markdown | 8.0 KB |
| frameworks/recruiter-playbook.md | markdown | 9.4 KB |
| prompts/counter-simulation.md | markdown | 12.9 KB |
| prompts/final-recommendation.md | markdown | 8.7 KB |
| prompts/leverage-map.md | markdown | 6.6 KB |
| prompts/negotiation-strategy.md | markdown | 7.5 KB |
| prompts/offer-diagnosis.md | markdown | 7.5 KB |
| prompts/script-generator.md | markdown | 11.0 KB |
frameworks/leverage-heuristics.md
View on GitHub# Leverage Heuristics
判定 offer 每个组件的 leverage tier — **LOW / MEDIUM / HIGH** — 的规则手册。
在 [`../prompts/offer-diagnosis.md`](../prompts/offer-diagnosis.md) Step 3 逐项打标签时按这个查。
---
## 通用规则
**Leverage = 招聘官在这一项上有多少空间 × 你的推动力**
3 个维度联合决定:
| 维度 | 高 leverage | 低 leverage |
|---|---|---|
| **市场位置** | Offer 低于 P50 | Offer 高于 P75 |
| **你手中的替代** | 有书面 competing offer / 现职稳定 | 无 offer / 待业 3+ 月 |
| **招聘官内部约束** | 该组件不在硬 band 里 | 该组件是 policy 死数 |
**3 个都高 = HIGH。1 个高 2 个低 = MEDIUM 或 LOW(视组件而定)。全低 = LOW。**
---
## 逐组件 leverage 判定表
### Base Salary
| 场景 | Leverage | 说明 |
|---|---|---|
| 低于 P50 + 有 competing offer | HIGH | 招聘官知道自己低,且怕失去你 |
| 低于 P50 + 无 competing offer | MEDIUM | 数据在你这边但没杠杆 |
| P50-P75 + 有 competing offer | MEDIUM | 主要靠 competing offer 换空间 |
| P50-P75 + 无 competing offer | LOW-MED | 只能靠 justification 强推 |
| 大于 P75 | LOW | 已到 band 顶,硬推=消耗信用 |
| Level 已 down-graded | LOW | 招聘官会说 "you're already at top of the lower band" |
| Recruiter 明确说 "base is capped for this level" | LOW(立刻) | 别再 push base,pivot 到 RSU |
**特殊:Netflix / 某些 hedge fund** — base 是唯一 comp。此时"base low leverage"规则不适用,base 就是主战场。
---
### RSU (Equity)
| 场景 | Leverage | 说明 |
|---|---|---|
| 4-yr avg 低于 P50 | HIGH | 且大多数公司 RSU approver ≠ base approver,弹性大 |
| Base 已 capped + RSU 未到 P75 | HIGH | Recruiter 天然会想"我从 RSU 补" |
| 私营公司 + illiquidity discount 未 factor | MEDIUM | 你可以 argue "at 30% discount, effective TC is X" |
| 公司股价近期跌 30%+ | MEDIUM | Recruiter 会说 "we're being conservative" 但 approver 有时给"股数补偿" |
| Grant 已 include refresh commitment | LOW | 前置 refresh = 招聘官已经 pre-loaded 你的 TC 预期 |
| 大于 P75 | LOW | 硬推 = "reconsider pool" 风险 |
**特殊:Amazon** — RSU 极 back-loaded(5/15/40/40),Y1/Y2 sign-on 通常是主战场,RSU 本身谈判空间较小(Amazon 的 comp model 依赖 back-loaded stock)。
---
### Sign-on Bonus
| 场景 | Leverage | 说明 |
|---|---|---|
| Base + RSU 都在 band cap | HIGH | Sign-on 是最后一块 dumping ground,approver 通常低(recruiter 有 discretion) |
| 用户放弃 unvested equity(真的) | HIGH | 最强天然理由,几乎所有大厂 accept |
| Relocation 未 cover 全部 | MEDIUM | "relocation actual cost exceeds lump sum by $X" 好使 |
| 已经给过一次 sign-on | LOW | 二次 push = greedy signal |
| 公司政策不给 sign-on(罕见) | 零 | Netflix 等 |
| Clawback 24 个月 + 你不确定长留 | LOW(你自己的判断) | 别硬要,clawback = 锁 |
**特殊:中国大厂** — sign-on 概念弱,谈的是"签字奖"和"补前公司未 vested 期权"。
---
### Annual Bonus
| 场景 | Leverage | 说明 |
|---|---|---|
| Year 1 not guaranteed(default) | MEDIUM | "Guarantee Year 1 at target" 是标准要求,大厂通常 accept |
| Bonus target % 谈判 | LOW(几乎不) | Target % 挂 Level,不谈;除非同时 push Level |
| Sign 到中年(如 8 月)Y1 pro-rated | HIGH | "Pro-rated will disadvantage me, please guarantee 100%" 好使 |
| 从年度 target 20% 改成 30% | 零 | Level 挂钩,不谈 |
**别浪费谈判信用在 bonus % 上。** 除非 Year 1 not guaranteed,几乎不值得单独提。
---
## 特殊 leverage 类型(非现金)
### Refresh Commitment
- 招聘官承诺 "you'll be eligible for refresh grant at year 1, targeting X% of initial grant"
- **不能进 offer letter**(policy 不允许),但可以进邮件 confirm
- 招聘官很爱给这个 —— 花不到他的现金预算
- **报告应该建议**:Modality A/B 谈判中,如果 recruiter 拒了 base / RSU 加价,主动要 "written refresh commitment in email confirmation"
### Early Performance Review
- "6-month review instead of 12-month"
- 加速 promotion / merit increase 节奏
- Recruiter discretion 内,通常给
- **报告应该建议**:非 competing offer 情况下的 golden request
### Start Date Flexibility
- 早 start = 你自己让步
- 晚 start(3-4 周内)= 你的杠杆(recruiter Q3 KPI 压力)
- 晚 start(8+ 周)= recruiter 通常拒(headcount 会被 clawback)
### Interview Feedback Loop
- 少见但强:让 recruiter 分享 "onsite feedback quotes"
- 有些 recruiter 会给部分原话,你可以拿来 anchor:"The panel loved my <specific area>, which is the exact skill this role needs — I think that value warrants closer to $X"
- **报告应该建议**:Modality A 谈判中的补充 justification
---
## 综合 verdict formula(Overall Leverage)
**每组件的 leverage tier 加权**:
- Base HIGH = 3, MED = 2, LOW = 1
- RSU HIGH = 3, MED = 2, LOW = 1
- Sign-on HIGH = 2, MED = 1.5, LOW = 1(权重稍低,因为它是补齐工具不是主战场)
- Bonus HIGH = 1, MED = 0.5, LOW = 0(几乎不影响 overall)
**总分**:
- **8+** → Overall HIGH → Modality A(双开激进)
- **6-7** → Overall MEDIUM → Modality B(单点集中)
- **4-5** → Overall LOW → Modality C(收尾轻抗)
- **3 or below** → Overall MINIMAL → Modality D(不谈,直接签)
**但**:**如果用户有书面 competing offer 且 competing offer TC 高 10%+**,总分 +2(自动升级至少一档)。
**但但**:**如果用户 待业 3+ 月 或 visa 60 天窗口**,总分 -2(自动降级至少一档,保守打)。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |