career-planner-china

原始内容

---
name: Career
slug: career
version: 1.0.5
description: >-
  Advises on career decisions — offer evaluation, salary and equity negotiation, promotions,
  pivots, and layoff response. Use when the user receives an offer or a recruiter ping, feels
  underpaid or stuck at a level, weighs staying vs quitting, gets passed over for promotion,
  faces a PIP, layoff, firing, or rescinded offer, compares equity and vesting packages, or
  plans a change of role, industry, or IC-to-manager track. Not for writing the resume itself
  (resume) or running the application pipeline (job-search).
homepage: https://clawic.com/skills/career
changelog: "Display name shown correctly"
metadata:
  clawdbot:
    emoji: "💼"
    displayName: Career
    configPaths:
    - ~/Clawic/data/career/
---

Advise mode: this skill guides a human making their own career moves; it never acts on their behalf. User career data lives in `~/Clawic/data/career/` (see `setup.md` on first use, `memory-template.md` for file formats): `profile.md` holds comp history, market band, constraints, and stated values — create it on the first substantive session and update after each decision. If you have data at an old location (`~/career/` or `~/clawic/career/`), move it to `~/Clawic/data/career/`.

## When To Use

- An offer or counteroffer is on the table and the user must respond
- Deciding whether to stay, leave, or pivot role, company, or track
- Building a promotion case, asking for a raise, or diagnosing a stalled level
- Responding to a layoff, PIP, firing, or rescinded offer
- Valuing equity, comparing offers, or benchmarking comp against the market
- Not for writing the resume itself (resume) or running the application pipeline (job-search)

## Quick Reference

| Situation | Play |
|---|---|
| Offer received | Never accept in the call; written details, 48 hours, counter once minimum (`offers.md`) |
| Asked for salary expectations first | Deflect once; if forced, state a range whose bottom is your target (`offers.md`) |
| Exploding deadline (under 1 week) | Request one week in writing; treat refusal as data about the employer |
| Recruiter ping while happy | Extract the band before committing to a process (`market.md`) |
| Offer heavy on equity or options | Risk-adjust before comparing — face value is the optimistic read (`equity.md`) |
| "Should I quit?" | No decision without a BATNA (Rule 1); run the razor (→ The Decision Razor) |
| Stalled 2+ promotion cycles | Diagnose scope vs sponsor vs structure (`promotion.md`) |
| Underpaid suspicion | Confirm with band data, then internal ask before external move (`market.md`) |
| Changing role and industry at once | One variable per move (Rule 4; `pivots.md`) |
| Laid off, fired, or PIP'd | 48-hour sequence; sign nothing on day one (`layoffs.md`) |
| Counteroffer after resigning | Decline by default: root cause persists and flight risk is now on record |
| Director-and-above move | Different mechanics: negotiate severance at entry (`executive.md`) |
| Happy, no active decision | Market check every 18-24 months: 2-3 recruiter conversations to reprice the band |
| Anything else | Collect three facts before advising: current total comp, market band, live alternatives |

Depth on demand: `offers.md` evaluate, compare, counter · `equity.md` RSUs, options, vesting math · `promotion.md` levels, calibration, internal raises · `pivots.md` role, industry, track changes, breaks, going independent · `layoffs.md` layoffs, PIPs, severance, visas · `market.md` benchmarking and staying liquid · `executive.md` director-and-above moves.

## Core Rules

1. **No BATNA, no decision.** "Stay or go" without a concrete outside option is a mood poll. Check: can the user name a specific alternative and its comp? If not, the first step is generating options, not deciding.
2. **Counter = 2 × target − offer.** Offer 100k, target 112k, counter 124k: negotiations gravitate toward the midpoint of the two stated numbers, so countering at your target guarantees landing below it.
3. **Switchers outprice stayers.** Atlanta Fed Wage Growth Tracker: job switchers beat stayers by roughly 1-2 percentage points of annual wage growth in most years. This is the base-rate argument behind the 18-24 month market check even when content.
4. **One-variable transitions.** Per move, change role OR industry OR track (IC to manager), not two. Each simultaneous change resets leverage to beginner on that axis; two at once typically costs both comp and title.
5. **Promotion is a lagging indicator.** You get promoted for having already done the next level's job for 6-12 months with witnesses outside your team. If nobody above your manager can describe your work, the packet fails at calibration.
6. **Negotiate the package, not the base.** Typical employer flexibility, most to least: sign-on bonus, equity, start date and vacation, title and review timing, base. Sign-on is a one-time cost to them; moving base reprices their whole band.
7. **Reversibility sets the risk bar.** If a move can be undone within about 12 months (boomerang hire, internal transfer back, contract work), bias toward acting; irreversible moves (relocation with family, visa changes, walking away from a cliff) get the full BATNA-plus-runway treatment.
8. **Resign only on paper.** A verbal offer is a mood; resign only on a written offer with contingencies (background check, references) cleared. A rescind after resignation is a self-inflicted layoff.

## The Decision Razor

Diagnose before prescribing; a bad manager and a bad career have opposite fixes.

- (a) Would a different manager fix this? Internal transfer before quitting (`promotion.md`). (b) Would the same role at a better company fix it? Switch companies, keep the role (Rule 4). (c) Does the work itself drain you even in good weeks, for 3+ months? Pivot, planned as a one-variable move (`pivots.md`).
- Quitting without an offer requires runway of at least 2× realistic search time. Senior searches commonly run 3-6 months, so hold 6-12 months of expenses in cash, not "some savings"; executive searches run longer — scale the same formula (`executive.md`).
- Tenure math: sunk years are not an asset; unvested equity is. Compute the dollar cost of leaving before the next vest date (`equity.md`) and treat only that number as the price of leaving.

## Offer Response In Brief

Full sequence in `offers.md`; the spine:

1. Everything in writing: base, bonus target with payout history, equity grant with vest schedule and cliff, sign-on with clawback terms, refresher policy.
2. Compute annualized total comp yourself; recruiters quote best case.
3. Set target from the market band (`market.md`), never from current comp. "Current plus 15%" is an anchor, not a target.
4. Counter once at 2 × target − offer (Rule 2), justified with one sentence of market data, never personal need.
5. Base capped? Walk the flexibility ladder in Rule 6 order.

Internal negotiation is a different game: performance already known, peer equity beats market rate, and you need a champion in the room, not just a request (`promotion.md`). Never wave an external offer you are unwilling to take.

## Output Gates

Before delivering career advice, check:

- Did I get current total comp, market band, and live alternatives before recommending any number or any stay/quit call?
- Is every counter I proposed computed as 2 × target − offer, not the midpoint of the visible range?
- Did I check unvested equity and cliff dates before endorsing a departure timeline?
- When no live alternative exists, did I recommend generating a BATNA instead of deciding?
- If `visa_dependent` or a non-US `region` applies, did I adjust legal defaults and timelines (`layoffs.md`)?
- Is the recommendation scored against the values in `profile.md`, not against brand prestige?

## Configuration

User-dependent variables. Defaults apply until the user states a preference; store them in `~/Clawic/data/career/config.yaml`.

| Variable | Type | Default | Effect |
|---|---|---|---|
| region | US \| EU \| UK \| other | US | Switches employment-law defaults in `layoffs.md`: at-will vs statutory notice, severance norms, non-compete enforceability |
| career_stage | early \| mid \| senior \| executive | mid | Routes director-and-above questions to `executive.md`; scales search-time and runway estimates in The Decision Razor |
| risk_posture | conservative \| moderate \| aggressive | moderate | Conservative doubles runway floors and defaults against irreversible moves (Rule 7); aggressive accepts reversible bets at 1× runway |
| visa_dependent | bool | false | Adds immigration clocks to every timeline: layoff grace periods, job-change risk, start-date and termination-date negotiation |

Preference areas to record as the user reveals them:

- **comp philosophy** — cash vs equity weighting, appetite for private-company paper; affects offer scoring in `offers.md` and equity discounting in `equity.md`
- **constraints** — location and remote requirements, excluded industries, family or caregiving limits; affects option generation before any stay/leave call
- **values ranking** — what "better" means (comp, learning, autonomy, mission, hours), stored ranked in `profile.md`; affects how offers and stay/leave options are scored
- **coaching style** — frameworks only vs drafted word-for-word scripts, blunt vs cushioned delivery; affects the format of every negotiation deliverable

## Traps

| Trap | Why it fails | Do instead |
|---|---|---|
| Accepting on the call | Peak-emotion decision at their anchor | 48 hours and written details, always |
| Naming your number first | Anchors at your guess of their band, usually low | Deflect once; if forced, range with target at the bottom |
| Comparing offers on year-1 base | Bonus reality, vest schedule, and refreshers dominate by year 3 | Normalize to 4-year total value (`offers.md`) |
| Resigning to force a counteroffer | Root cause persists and you are flagged as flight risk | Resign only when ready to leave |
| "I have invested 5 years here" | Sunk cost; tenure is not an asset, unvested equity is | Price only the forward-looking cost of leaving |
| Chasing title over scope | Inflated title with unchanged scope fails the next interview loop | Trade title for scope, budget, or reports |
| Calling a mentor a sponsor | Advice does not move calibration decisions | Verify who speaks for you in the room (`promotion.md`) |
| Negotiating base only | Base is their stickiest number | Sign-on, equity, dates, title first (Rule 6) |
| Interviewing only when desperate | Desperation prices in; the band is stale exactly when you need it | Market check on elapsed time, not unhappiness (`market.md`) |
| Prestige capture | Optimizing for logo approval over fit compounds misery | Score offers against stored profile values, not brand names |

## Where Experts Disagree

- **Passion-first vs skill-first**: both schools hold conditionally. Under about 5 years of experience, skills compound faster than self-knowledge, so bias skill-first (the career-capital position, Newport); with rare skills banked, passion picks between good options.
- **Generalist vs specialist**: specialize while the field expands (early markets pay depth); generalize as it consolidates (mature markets pay range and management).
- **Loyalty premium**: tenure-track systems (government, academia, some East Asian conglomerates) genuinely pay tenure; the switcher math of Rule 3 is calibrated to open-market private sectors and reverses there.

## Related Skills
More Clawic skills, get them at https://clawic.com/skills/career (install if the user confirms):
- `job-search` — running the pipeline: applications, company research, interview prep
- `resume` — when the decision is made and materials need targeting
- `negotiate` — negotiation mechanics beyond comp: limits, principals, approvals
- `comp-design` — the employer side: designing base, bonus, and equity mix

## Feedback

- If useful, star it: https://clawic.com/skills/career
- Latest version: https://clawic.com/skills/career

Part of [Clawic](https://clawic.com), the verified skill library. Get this skill: https://clawic.com/skills/career.

name: Career slug: career version: 1.0.5 description: >- Advises on career decisions — offer evaluation, salary and equity negotiation, promotions, pivots, and layoff response. Use when the user receives an offer or a recruiter ping, feels underpaid or stuck at a level, weighs staying vs quitting, gets passed over for promotion, faces a PIP, layoff, firing, or rescinded offer, compares equity and vesting packages, or plans a change of role, industry, or IC-to-manager track. Not for writing the resume itself (resume) or running the application pipeline (job-search). homepage: https://clawic.com/skills/career changelog: "Display name shown correctly" metadata: clawdbot: emoji: "💼" displayName: Career configPaths: - ~/Clawic/data/career/

Advise mode: this skill guides a human making their own career moves; it never acts on their behalf. User career data lives in ~/Clawic/data/career/ (see setup.md on first use, memory-template.md for file formats): profile.md holds comp history, market band, constraints, and stated values — create it on the first substantive session and update after each decision. If you have data at an old location (~/career/ or ~/clawic/career/), move it to ~/Clawic/data/career/.

When To Use

An offer or counteroffer is on the table and the user must respond
Deciding whether to stay, leave, or pivot role, company, or track
Building a promotion case, asking for a raise, or diagnosing a stalled level
Responding to a layoff, PIP, firing, or rescinded offer
Valuing equity, comparing offers, or benchmarking comp against the market
Not for writing the resume itself (resume) or running the application pipeline (job-search)

Quick Reference

Situation	Play
Offer received	Never accept in the call; written details, 48 hours, counter once minimum (`offers.md`)
Asked for salary expectations first	Deflect once; if forced, state a range whose bottom is your target (`offers.md`)
Exploding deadline (under 1 week)	Request one week in writing; treat refusal as data about the employer
Recruiter ping while happy	Extract the band before committing to a process (`market.md`)
Offer heavy on equity or options	Risk-adjust before comparing — face value is the optimistic read (`equity.md`)
"Should I quit?"	No decision without a BATNA (Rule 1); run the razor (→ The Decision Razor)
Stalled 2+ promotion cycles	Diagnose scope vs sponsor vs structure (`promotion.md`)
Underpaid suspicion	Confirm with band data, then internal ask before external move (`market.md`)
Changing role and industry at once	One variable per move (Rule 4; `pivots.md`)
Laid off, fired, or PIP'd	48-hour sequence; sign nothing on day one (`layoffs.md`)
Counteroffer after resigning	Decline by default: root cause persists and flight risk is now on record
Director-and-above move	Different mechanics: negotiate severance at entry (`executive.md`)
Happy, no active decision	Market check every 18-24 months: 2-3 recruiter conversations to reprice the band
Anything else	Collect three facts before advising: current total comp, market band, live alternatives

Depth on demand: offers.md evaluate, compare, counter · equity.md RSUs, options, vesting math · promotion.md levels, calibration, internal raises · pivots.md role, industry, track changes, breaks, going independent · layoffs.md layoffs, PIPs, severance, visas · market.md benchmarking and staying liquid · executive.md director-and-above moves.

Core Rules

No BATNA, no decision. "Stay or go" without a concrete outside option is a mood poll. Check: can the user name a specific alternative and its comp? If not, the first step is generating options, not deciding.
Counter = 2 × target − offer. Offer 100k, target 112k, counter 124k: negotiations gravitate toward the midpoint of the two stated numbers, so countering at your target guarantees landing below it.
Switchers outprice stayers. Atlanta Fed Wage Growth Tracker: job switchers beat stayers by roughly 1-2 percentage points of annual wage growth in most years. This is the base-rate argument behind the 18-24 month market check even when content.
One-variable transitions. Per move, change role OR industry OR track (IC to manager), not two. Each simultaneous change resets leverage to beginner on that axis; two at once typically costs both comp and title.
Promotion is a lagging indicator. You get promoted for having already done the next level's job for 6-12 months with witnesses outside your team. If nobody above your manager can describe your work, the packet fails at calibration.
Negotiate the package, not the base. Typical employer flexibility, most to least: sign-on bonus, equity, start date and vacation, title and review timing, base. Sign-on is a one-time cost to them; moving base reprices their whole band.
Reversibility sets the risk bar. If a move can be undone within about 12 months (boomerang hire, internal transfer back, contract work), bias toward acting; irreversible moves (relocation with family, visa changes, walking away from a cliff) get the full BATNA-plus-runway treatment.
Resign only on paper. A verbal offer is a mood; resign only on a written offer with contingencies (background check, references) cleared. A rescind after resignation is a self-inflicted layoff.

The Decision Razor

Diagnose before prescribing; a bad manager and a bad career have opposite fixes.

(a) Would a different manager fix this? Internal transfer before quitting (promotion.md). (b) Would the same role at a better company fix it? Switch companies, keep the role (Rule 4). (c) Does the work itself drain you even in good weeks, for 3+ months? Pivot, planned as a one-variable move (pivots.md).
Quitting without an offer requires runway of at least 2× realistic search time. Senior searches commonly run 3-6 months, so hold 6-12 months of expenses in cash, not "some savings"; executive searches run longer — scale the same formula (executive.md).
Tenure math: sunk years are not an asset; unvested equity is. Compute the dollar cost of leaving before the next vest date (equity.md) and treat only that number as the price of leaving.

Offer Response In Brief

Full sequence in offers.md; the spine:

Everything in writing: base, bonus target with payout history, equity grant with vest schedule and cliff, sign-on with clawback terms, refresher policy.
Compute annualized total comp yourself; recruiters quote best case.
Set target from the market band (market.md), never from current comp. "Current plus 15%" is an anchor, not a target.
Counter once at 2 × target − offer (Rule 2), justified with one sentence of market data, never personal need.
Base capped? Walk the flexibility ladder in Rule 6 order.

Internal negotiation is a different game: performance already known, peer equity beats market rate, and you need a champion in the room, not just a request (promotion.md). Never wave an external offer you are unwilling to take.

Output Gates

Before delivering career advice, check:

Did I get current total comp, market band, and live alternatives before recommending any number or any stay/quit call?
Is every counter I proposed computed as 2 × target − offer, not the midpoint of the visible range?
Did I check unvested equity and cliff dates before endorsing a departure timeline?
When no live alternative exists, did I recommend generating a BATNA instead of deciding?
If visa_dependent or a non-US region applies, did I adjust legal defaults and timelines (layoffs.md)?
Is the recommendation scored against the values in profile.md, not against brand prestige?

Configuration

User-dependent variables. Defaults apply until the user states a preference; store them in ~/Clawic/data/career/config.yaml.

Variable	Type	Default	Effect
region	US \| EU \| UK \| other	US	Switches employment-law defaults in `layoffs.md`: at-will vs statutory notice, severance norms, non-compete enforceability
career_stage	early \| mid \| senior \| executive	mid	Routes director-and-above questions to `executive.md`; scales search-time and runway estimates in The Decision Razor
risk_posture	conservative \| moderate \| aggressive	moderate	Conservative doubles runway floors and defaults against irreversible moves (Rule 7); aggressive accepts reversible bets at 1× runway
visa_dependent	bool	false	Adds immigration clocks to every timeline: layoff grace periods, job-change risk, start-date and termination-date negotiation

Preference areas to record as the user reveals them:

comp philosophy — cash vs equity weighting, appetite for private-company paper; affects offer scoring in offers.md and equity discounting in equity.md
constraints — location and remote requirements, excluded industries, family or caregiving limits; affects option generation before any stay/leave call
values ranking — what "better" means (comp, learning, autonomy, mission, hours), stored ranked in profile.md; affects how offers and stay/leave options are scored
coaching style — frameworks only vs drafted word-for-word scripts, blunt vs cushioned delivery; affects the format of every negotiation deliverable

Traps

Trap	Why it fails	Do instead
Accepting on the call	Peak-emotion decision at their anchor	48 hours and written details, always
Naming your number first	Anchors at your guess of their band, usually low	Deflect once; if forced, range with target at the bottom
Comparing offers on year-1 base	Bonus reality, vest schedule, and refreshers dominate by year 3	Normalize to 4-year total value (`offers.md`)
Resigning to force a counteroffer	Root cause persists and you are flagged as flight risk	Resign only when ready to leave
"I have invested 5 years here"	Sunk cost; tenure is not an asset, unvested equity is	Price only the forward-looking cost of leaving
Chasing title over scope	Inflated title with unchanged scope fails the next interview loop	Trade title for scope, budget, or reports
Calling a mentor a sponsor	Advice does not move calibration decisions	Verify who speaks for you in the room (`promotion.md`)
Negotiating base only	Base is their stickiest number	Sign-on, equity, dates, title first (Rule 6)
Interviewing only when desperate	Desperation prices in; the band is stale exactly when you need it	Market check on elapsed time, not unhappiness (`market.md`)
Prestige capture	Optimizing for logo approval over fit compounds misery	Score offers against stored profile values, not brand names

Where Experts Disagree

Passion-first vs skill-first: both schools hold conditionally. Under about 5 years of experience, skills compound faster than self-knowledge, so bias skill-first (the career-capital position, Newport); with rare skills banked, passion picks between good options.
Generalist vs specialist: specialize while the field expands (early markets pay depth); generalize as it consolidates (mature markets pay range and management).
Loyalty premium: tenure-track systems (government, academia, some East Asian conglomerates) genuinely pay tenure; the switcher math of Rule 3 is calibrated to open-market private sectors and reverses there.

Related Skills

More Clawic skills, get them at https://clawic.com/skills/career (install if the user confirms):

job-search — running the pipeline: applications, company research, interview prep
resume — when the decision is made and materials need targeting
negotiate — negotiation mechanics beyond comp: limits, principals, approvals
comp-design — the employer side: designing base, bonus, and equity mix

Feedback

If useful, star it: https://clawic.com/skills/career
Latest version: https://clawic.com/skills/career

Part of Clawic, the verified skill library. Get this skill: https://clawic.com/skills/career.