# The Solo-Founder Cofounder Slate: Missing Role to Met, Cleared Candidates

*Starting from one solo founder and one missing role, produce in two weeks a slate of 8 to 12 named, conflict-checked candidates the founder has met, plus an equity position.*

- Canonical URL: https://www.refolk.ai/guides/solo-founder-cofounder-slate-playbook
- Pillar: Investing and deal sourcing
- Format: Playbook
- Published: 2026-11-12
- Last reviewed: 2026-11-12
- Reading time: 22 min
- Keywords: how to find a technical cofounder for a portfolio company, pre-seed solo founder cofounder search, venture studio cofounder matching process, late-joining cofounder equity and vesting, help solo founder find cofounder investor platform team

## Key takeaways

- Solo founders were roughly 35% of startups incorporated on Carta but only about 17% of those that closed a venture round within the year, so a cofounder search is a de-risking event the next round will price.
- In Refolk's index, 9,686 US profiles carry a Founder or Co-Founder title with "stealth" in the headline, 2.16 times the 4,487 US Founding Engineers, so many ready candidates are already founders who will expect cofounder-level equity.
- A stealth headline is noisy: in Refolk's samples of 25, 3 US and 2 UK results were investors, academics or trustees rather than operators, so require a full-time current role that started within the last 12 months.
- A YC match only means two people agreed to talk; YC's 2021 figures show 130,000 invites at 25% acceptance producing 33,000 matches, and YC has not published how many became companies, so count only dated, logged founder meetings.
- Late-joiner equity has no published benchmark: the anchors run from a 51-49 median split for founders who started together to a 1.49% median first-hire grant, about a 33x spread, so build the starting position from months since founding and traction.
- Clear covenants before every introduction: California bars post-employment non-competes, but employee non-solicits are unsettled there and enforced in other states, and side-project code escapes assignment only if built on own time with no employer equipment.

This playbook is for the platform partner, talent partner or angel whose solo founder needs a technical or commercial cofounder. It turns "you need a partner" into a bounded two-week search that ends in a slate of 8 to 12 named candidates, each with a written gap-fit rationale, a dated readiness cue, a portfolio and covenant conflict check, and a logged first meeting with the founder, plus a written equity and vesting starting position for the conversations that follow.

## What a cofounder slate is, and why the fund runs the search

A cofounder slate is a short list of 8 to 12 people who could credibly fill the founder's missing role, have a dated reason to be available now, are clear of portfolio and covenant conflicts, and have already met the founder. The fund runs the search because a funded solo founder is an exception the next round will price, and a bounded search from the investor side closes that risk faster than open matching.

Four terms carry the method:

- **Gap-fit rationale**: two or three written sentences tying a candidate's track record to the specific weaknesses in the founder's gap brief, not to the role title.
- **Readiness cue**: a dated, individual event suggesting this person may start or join something now, such as a departure, a new stealth title, or own-time building.
- **Conflict check**: the screen for portfolio poaching, restrictive covenants and invention-assignment claims, run before the founder meets anyone.
- **Starting position**: the written equity range and vesting terms the founder brings into the first serious conversation. It is not an offer.

The numbers explain why investors care. Carta's Solo Founders Report found solo founders made up 36.3% of new startups in its most recent half-year, up from 23.7% six years earlier. Funding does not follow the headcount. Solo founders were roughly 35% of startups incorporated but only about 17% of those that closed a venture round within the year, and solo-led companies took 14.7% of the cash raised in priced equity rounds while making up 30% of startups founded. Y Combinator reports that only four of its top 100 companies came to it without a cofounder.

These are mostly US figures from one cap-table platform, and nobody has published the solo share at the first priced round. The direction is still clear: there is roughly a 2x gap between solo founders' share of companies and their share of rounds, which is key-person risk being priced. When you back a solo founder, the cofounder search is the de-risking event your next co-investor will ask about.

One more number shapes the first conversation. Solo founders bring on their first employees earlier than teams, in about 399 days on average against 480. Your founder may already have a strong first engineer or seller, so the gap brief has to decide honestly whether the missing role is a cofounder or a senior hire. Running a cofounder search for what is really a senior hire wastes two weeks and a lot of equity.

## How long a cofounder search takes, and why two weeks is enough

Open cofounder matching usually takes months; a fund-run search can produce a met slate in two weeks because it starts from a dated, pre-qualified pool instead of waiting for strangers to find each other. The two weeks produce a slate of met candidates, not a signed cofounder. Trial work runs two to six weeks after that.

The published benchmarks come from matching programmes and accelerators:

| Programme | Metric | Value |
|---|---|---|
| YC Co-Founder Matching | Invite acceptance | 25% |
| YC Co-Founder Matching | Matches per person | 3.6 |
| YC (secondary, unverified) | Median days, signup to eventual cofounder | About 100 |
| Antler London (one cohort) | Teams formed that reached IC | 13 of 49 (26.5%) |
| Antler | Team formation window | Weeks 2 to 4 |

Read each row for what it proves:

- **YC's figures are about introductions, not outcomes.** In its 2021 disclosure, founders sent 130,000 invites, a quarter were accepted, and the platform made 33,000 matches. YC has not published how many matches turned into companies.
- **The 100-day median is soft.** It comes from a secondary source, which also says some pairs took over nine months. Use it as an order of magnitude, not a target.
- **The Antler figure is an anecdote.** It is one participant's account of one cohort. Only teams that pass the investment committee (IC) are funded; in the US residency that is $250K at a $2.75M post-money valuation.

Your two weeks are shorter than any of these benchmarks. That works only because of three differences from open matching: you already know the founder's gap, you source people who carry dated cues rather than an open sign-up pool, and you control the meeting calendar. Drop any one of the three and you are back on the open-matching clock. Venture studios and residencies solve the same problem by putting candidates in one room for weeks; a fund platform team substitutes preparation for proximity.

## Where ready candidates sit: the pre-founding signal pools

Ready candidates cluster in visible pools: people with a stealth founder title, founding engineers at early companies, and recent leavers from venture-backed employers. Each pool tells you where to look, but membership counts as readiness only when it is paired with a dated cue about the individual.

Refolk's index sizes the first two pools directly. The sample column comes from samples of 25 profiles per segment.

| Segment | Count | Stealth-named employers in sample | Multiple vs UK stealth |
|---|---|---|---|
| US Founder/Co-Founder + "stealth" | 9,686 | 7/25 (28%) | 4.35x |
| UK Founder/Co-Founder + "stealth" | 2,229 | 9/25 (36%) | 1.0x |
| US Founding Engineer | 4,487 | 0/25 | 2.01x |

**9,686 - US profiles with a Founder or Co-Founder title and "stealth" in the headline**

Counted in Refolk's index of professional profiles; the same query for the United Kingdom returns 2,229.

Three readings follow from the table:

1. **Stealth founders are founders first.** In Refolk's index, US stealth founders outnumber US founding engineers 2.16x. Many people labelled stealth have already started something and are looking for a partner, not a seat. Expect them to want cofounder-level equity, not a first-hire grant, and expect a few to pitch your founder on joining them instead.
2. **The UK pool is smaller but more explicit.** It is about a quarter the size of the US pool (4.35x smaller), yet stealth-named employers appear in 36% of the UK sample against 28% in the US, so UK headlines say what they mean slightly more often.
3. **Founding engineers are spread thin.** In the US Founding Engineer sample, each of the top 10 employers appears exactly once. There is no single company to mine, and each person is probably load-bearing at an early company, which raises both the poaching question and the readiness bar.

Demand matters as much as supply. On YC's matching platform, 63% of founders wanted a cofounder who does engineering, so a technical search competes with most other searchers for the same people. A commercial search draws on a less contested queue, though no published figure sizes that pool.

The third pool is defined by employer. Gompers, Lerner and Scharfstein found in the Journal of Finance that, controlling for firm size, patents and industry, the most prolific spawners of new companies are originally venture-backed firms in Silicon Valley and Massachusetts, and that when growth slows, employees are likely to leave to start new ones. That is a finding about firms. Use it to choose where to look, never to decide who is ready.

### Readiness cues: what each proves and how it lies

| Cue | What it proves | When it lies | Check |
|---|---|---|---|
| Stealth founder title | Has started, may want a partner | Investors and academics list it as a side interest | Full-time role started in last 12 months |
| Venture-backed employer | Works where founders spawn | Says nothing about this person | Pair with an individual dated cue |
| Employer growth slowing | Firm-level push to leave | Applies to everyone there | Pair with an individual dated cue |
| Own-time repo activity | Is building something | Code written on employer kit | Ask what equipment, and when |
| Founding Engineer title | Has built from zero before | Already committed to another early bet | Ask what would make them leave |

No published base rate exists for vest-cliff timing, stealth titles, advisor-only roles or side-project repo activity. Every cue in the table is a reason to call, not a prediction, and each one needs a date and a source logged next to it.

#### Which longlist names to call first

Horizontal axis runs from No dated individual cue to Dated individual cue. Vertical axis runs from Weak gap fit to Strong gap fit.

| Quadrant | What it means |
| --- | --- |
| Strong fit, no cue | Keep warm, ask about timing, do not count toward the slate |
| Strong fit, dated cue | Screen this week and prioritise for founder meetings |
| Weak fit, no cue | Drop from the longlist |
| Weak fit, dated cue | Do not let urgency fake fit; refer to another search |

*Gap fit decides whether a person belongs on the slate; a dated individual cue decides whether to call them this week.*

The slow part of step 3 is attaching a date and a source to every name across several places at once. In [Refolk](/) you can ask for the pool in plain English with the cue built into the question, and get people back from LinkedIn, GitHub and the open web without writing a separate string for each source.

I ran this search: `Engineers in the San Francisco Bay Area who left a Series B or later startup in the last six months and now list stealth or advisor roles` - [see the full result list](https://www.refolk.ai/s/w59esbrbp7).

*Returns named people with their previous employer and current stealth or advisor role, so each arrives with a departure to date-check and log as a readiness cue.*

## The two-week procedure, step by step

The procedure runs from a written gap brief on day 1 to a met, conflict-cleared slate with an equity starting position on day 14, with trial work scoped afterwards. Steps overlap on purpose: screening starts before the longlist is finished, and covenant checks run per candidate rather than in one batch.

#### Solo founder to met cofounder slate in 14 days

1. **Define the gap with the founder** - Days 1 and 2. Write the founder's real weaknesses, the stage, and the exact scope of the missing role, including whether it is a cofounder or a senior hire; done is a one-page gap brief with no blank lines.
2. **Set the conflict rules** - Day 2. The platform team and counsel list off-limits portfolio companies and agree how states of employment will be handled; done is a written exclusion list agreed before sourcing.
3. **Build a dated longlist of 30 or more** - Days 2 to 6. Source at least 30 names and log a dated readiness cue and its source for each; done is a longlist where every name carries a date.
4. **Run outreach and screening calls** - Days 4 to 9. Contact the longlist and screen against the gap brief; done is 8 to 12 people who want to go further, each with a written gap-fit rationale.
5. **Clear covenants and IP before each intro** - Rolling. Each candidate self-discloses their agreements and state of employment, and counsel reviews anything flagged; done is a cleared or flagged status per name.
6. **Log a founder meeting for every name** - Days 7 to 14. The founder meets each cleared candidate and the platform team logs date, format and both sides' reads; done is a dated log for all 8 to 12.
7. **Write the equity and vesting starting position** - Days 10 to 14. Build a range from months since founding, traction and role scope, plus vesting with a stated start date; done is a written range and terms.
8. **Scope trial work with a decision date** - After the slate. Scope a two- to six-week project with the strongest one or two candidates; done is a written scope and a fixed decision date.

### Step 1: the gap brief

YC's own advice to founders is to state your stage and the exact skill gap, not a generic "co-founder wanted." The same discipline applies when you write it on the founder's behalf. The hard line is "what the founder does badly or avoids": it must be backed by evidence from the board pack, customer calls or the codebase, not the founder's self-description.

**One-page gap brief**

```
Company:
Stage and capital raised to date:
Months since founding:
Traction already built (users, revenue, product shipped):
What the founder does well (evidence):
What the founder does badly or avoids (evidence, not self-description):
The missing role in one sentence:
Cofounder or senior hire, and why:
Must-have skills (three at most):
Nice-to-have skills:
Location and working pattern:
What this person owns in the first 12 months:
Rough equity and salary the founder can offer:
Deal-breakers on either side:
```

*Fill it with the founder in one sitting; if any line stays blank, the search is not ready to start.*

### Steps 3 and 4: longlist to screened

Thirty names is a floor. Screening calls are where the gap-fit rationale gets written, while the call is fresh. Ask every candidate the same three questions: what they would want to own, what would make them leave their current role, and what they need financially for the next 12 months.

### Step 6: the meetings

Move quickly from screen to founder meeting. A secondary source on YC's data reports that most accepted matches met within a day. Momentum is part of the signal, so book the founder meeting at the end of the screening call when you can.

## Clearing portfolio, covenant and IP conflicts before each introduction

Before any candidate meets the founder, check four layers in order: the fund's portfolio exclusion list, the candidate's state of employment, their own disclosure of their employment agreements, and the provenance of any side-project work. No regulator or industry body has documented a procedure for a fund to clear these before an introduction, so this is your fund's practice to own with counsel.

#### Conflict layers, outermost first

1. **Fund portfolio** - Off-limits companies on the exclusion list, checked first
2. **State of employment** - Non-compete and non-solicit rules differ by state
3. **Employment agreements** - The candidate's self-disclosed PIIA, confidentiality and non-solicit terms
4. **Side-project IP** - Built on own time, with no employer equipment or trade secrets

*A candidate is cleared only when every layer is checked, starting with the fund's own portfolio and ending with who owns their side project.*

A PIIA is a proprietary information and inventions assignment agreement: the document in which an employee assigns work-related inventions to their employer and agrees to keep its information confidential. What the law settles, and what it leaves open:

- **Non-competes in California.** Courts have interpreted Section 16600 to prohibit post-employment non-competes and post-employment customer non-solicits. SB 699 added § 16600.5, which makes it a civil violation for any employer to attempt to enter into or enforce a non-compete against a current, former or prospective California employee.
- **Employee non-solicits.** The California Supreme Court has not definitively addressed them, and other states enforce them. This matters if the candidate would bring former colleagues along.
- **Invention carve-outs.** California, Delaware, Illinois, Minnesota, North Carolina, Washington and Nevada have statutes limiting what an employer can claim. California's Labor Code 2870 covers inventions developed entirely on the employee's own time, with no equipment, supplies, facility or trade secret information of the employer.

Ask every candidate these before the introduction, and record the answers:

1. Which state are you employed in, and which state is your employer's agreement governed by?
2. Have you signed a PIIA, non-solicit or confidentiality agreement? Will you share it with counsel if needed?
3. Is any side project you would bring built on your own time and your own equipment?
4. Would you expect anyone from your current team to follow you?

> **Rule:** No introduction before a status
>
> Every candidate carries a recorded cleared or flagged status before the founder meets them. A flagged status goes to counsel, not onto the founder's calendar.

This section describes a process, not legal advice. When anything is flagged, counsel decides.

## Setting the equity and vesting starting position for a late-joining cofounder

The starting position is a written equity range and vesting terms built from first principles, because no published dataset breaks out cofounders who joined after founding. The two documented anchors sit about 33x apart, so the work is placing this candidate between them and writing down why.

| Position | Value |
|---|---|
| Two founders, median split | 51-49 |
| Two founders splitting exactly equally | 44.6% |
| Median first hire grant (four-year) | 1.49% |
| Founding team ownership at seed | 56.2% |

The top anchor describes two-founder teams that formed together, as reported by Carta. The bottom anchor is the median first hire at a US startup, on a four-year grant. The often-quoted "10 to 30%" for late joiners is a practitioner opinion with no data behind it. Quote it as opinion if you quote it at all.

#### Building the starting position

1. **Anchor the ends** - 51-49 for founders who started together, 1.49% for a median first hire
2. **Date the gap** - Write months since founding and traction already built
3. **Price the role** - Scope owned, salary given up, risk taken on
4. **Write a range** - A starting range with its reasoning, not an offer
5. **Set vesting** - Four years, one-year cliff, stated start date

*A late joiner's range is placed between two documented anchors using dated facts about the company, then wrapped in standard vesting.*

Rules for the written position:

- **Write the inputs next to the range.** Months since founding, capital raised and traction already built are what move a candidate down from the top anchor. Without them the range is a guess the candidate can argue in either direction.
- **Use four-year vesting with a one-year cliff.** That is documented as standard. Between 25% and 35% of two-founder teams founded over a recent five-year span had parted ways within five years, so vesting is how the founder survives a split.
- **State the vesting start date.** Nothing has been published on vesting start dates for late joiners. Pick one, usually the joining date or the trial start, and write it down so it is not negotiated by accident.
- **Leave acceleration to counsel.** Acceleration percentages attributed to Carta appear only on secondary sites. Treat them as unverified.
- **Remember the shrinking pie.** The median founding team owns 56.2% at seed, 36.1% at Series A and 23% at Series B. Equity granted before the next round is worth more to the candidate, which is leverage for the founder to close early.

If the candidate came from the stealth-founder pool, expect them to start near the top anchor. They are giving up their own company, not a job.

## How this goes wrong: failure modes and false positives

Most failed cofounder slates fail the same way: a signal that describes a pool, a platform or a different situation gets read as a fact about this candidate. Each failure below names the false positive it produces and the check that catches it.

1. **A stealth headline is noisy.** In Refolk's index samples, the stealth query also returned a General Partner, a Partner, an orthopaedics professor and a charity trustee: 3 of 25 in the US and 2 of 25 in the UK are not operators. False positive: investors and academics who list stealth as a side interest. Check: their current role is full-time and started within the last 12 months.
2. **Treating a match as a meeting.** A YC match only means both people agreed to talk. False positive: a slate counted in accepted invites rather than meetings held. Check: a dated, logged founder meeting for every name.
3. **Reading an employer signal as an individual one.** Gompers and colleagues measured which firms spawn founders, not which people will leave. False positive: everyone at a venture-backed company gets read as ready. Check: pair the employer signal with an individual dated cue.
4. **The California assumption.** Employee non-solicits are not settled even in California, and other states enforce them. False positive: a candidate cleared because "non-competes are unenforceable". Check: the candidate's state, plus their own PIIA disclosure, before the intro.
5. **Assuming the carve-out proves itself.** Side-project code written on an employer's laptop fails the own-time test in Labor Code 2870. False positive: a candidate whose project looks independent but is legally their employer's. Check: ask what equipment the work was done on, and when.
6. **Copying the equal-split anchor.** The 51-49 median describes founders who started together. False positive: a late joiner offered 50% because "that is the median". Check: write down months since founding and the traction already built before any number.
7. **Survivorship in programme figures.** The 13 of 49 Antler figure is one participant's cohort, and YC's 100-day median is secondary. False positive: a memo to the founder that promises an outcome rate. Check: label both as anecdote or unverified wherever they appear.

> **Watch out:** Urgency fakes fit
>
> A candidate with a fresh, dated cue and a weak gap fit is the most tempting mistake on any longlist. Readiness tells you when to call; only the gap brief tells you whether to.

> A match is an agreement to talk. A slate entry is a meeting that happened, with a date on it.

## Staging trial work after the slate

Trial work is a scoped, time-boxed project the founder and one or two candidates do together after the slate is delivered, with a decision date fixed in advance. Documented formats run from two to seven weeks, and none of them substitutes for the met slate that comes first.

What the sources document:

- **YC's Startup School** recommends "starting with 2-4 weeks of close work."
- **Antler London**, in one participant's account, ran a two-week bootcamp with "co-founder speed dating", then weeks three to seven in teams of two or three with a coach.
- **One VC blog** recommends a minimum trial period of four to six weeks.

Paid sprints and structured reference swaps have no studio or accelerator documentation, so their benefits are not established publicly. If you use them, treat the design as your own and review it after each search.

Sources also disagree on sequence. The VC blog puts the trial before any equity discussion; Antler forms teams first and runs the trial after. My position: always write the starting position in step 7 so the founder never improvises, then share it before the trial if the candidate is giving up salary or serious time, and after it if the trial is light.

Two rules make the trial safe. First, the project has a written scope and a decision date, so it ends in a yes or a no rather than drifting. Second, if the candidate is still employed, the trial runs on their own time and their own equipment, for the same Labor Code 2870 reasons that apply to side projects.

## Handing over the slate and keeping it current

The slate is done when every name has all five elements recorded in one place: gap-fit rationale, dated readiness cue, conflict status, logged founder meeting, and a shared equity starting position. After handover, the dated cues start to age, so the slate needs a re-check date of its own.

**Slate entry, one per candidate**

```
Name:
Current role, employer and start date:
Readiness cue (what, date, source):
Gap-fit rationale (two or three sentences, tied to the gap brief):
State of employment:
Portfolio conflict (none, or which company):
Covenant status (cleared or flagged, date, by whom):
Founder meeting (date, format, logged by):
Founder's read after the meeting:
Candidate's read after the meeting:
Next step and date:
```

*One block per name; the slate is not finished while any line is empty.*

#### Before you call the slate done

- [ ] The gap brief names the missing role in one sentence and says whether it is a cofounder or a senior hire.
- [ ] The portfolio exclusion list was agreed with counsel before sourcing started.
- [ ] The longlist reached at least 30 names, each with a dated readiness cue and a source.
- [ ] Every stealth-titled candidate holds a full-time current role that started within the last 12 months.
- [ ] Every employer-level signal is paired with an individual dated cue.
- [ ] Every slate name has a cleared or flagged covenant status recorded before their introduction.
- [ ] Every slate name has a dated, logged founder meeting, and none is counted from an accepted invite alone.
- [ ] The equity starting position is written as a range with months since founding and traction stated beside it.
- [ ] Vesting terms state four years, a one-year cliff, and an explicit start date.
- [ ] Any single-cohort or secondary figure in the memo to the founder is labelled as anecdote or unverified.
- [ ] Each trial project has a written scope and a fixed decision date.

What to do next:

- **Set a re-check date.** If no trial converts within the window, re-date every cue on the slate before reusing it. A departure from six months ago means something different a quarter later.
- **Rerun the same question.** Keep the plain-English query that built the longlist. Rerunning it in Refolk returns people whose cues are new since the last pass, so the second search starts from a fresh pool instead of a stale spreadsheet.
- **Re-check the legal layer by mechanism.** Covenant law changes by statute and by court decision. Before each new search, ask counsel whether the non-compete, non-solicit or carve-out position has moved in the candidate's state.
- **Feed the outcome back.** Record which pool and which cue produced the cofounder who signed. Over a few searches, that log becomes your fund's own base rate, which is the one figure nobody has published.

## Frequently asked questions

### How do I find a technical cofounder for a portfolio company?

Start from the founder's real weaknesses, not the job title, then source from dated pools rather than open matching sites. Write a one-page gap brief, set a portfolio exclusion list, build a longlist of 30 or more people who each carry a dated readiness cue, screen down to 8 to 12, clear covenants, and have the founder meet every one. Technical searches are crowded: 63% of founders on YC's matching platform wanted an engineering cofounder.

### How much equity should a late-joining cofounder get?

There is no published benchmark, so build it rather than copy it. The two documented anchors are a 51-49 median split for two founders who started together and a 1.49% median four-year grant for a first hire, about a 33x spread. The often-quoted 10 to 30% range is practitioner opinion. Place the candidate between the anchors using months since founding, traction already built, and the scope and risk of the role.

### How long does a pre-seed solo founder cofounder search take?

A fund-run search can produce a met slate of 8 to 12 candidates in two weeks, followed by two to six weeks of trial work. Open matching runs longer: a secondary, unverified source puts YC's median from signup to eventual cofounder at about 100 days. The two-week pace depends on starting from a known gap and a dated candidate pool, and on the platform team controlling the meeting calendar.

### Can I introduce employees of my own portfolio companies as cofounder candidates?

Treat them as off-limits unless that portfolio company's leadership agrees in advance. Poaching from your own portfolio damages the relationship you are paid to protect and can create a conflict for the fund. Write the exclusion list with counsel on day 2, before sourcing, so nobody spends a screening call on a candidate you cannot introduce. No public template for this list exists; it is each fund's own policy.

### Do non-solicit agreements stop a candidate from joining a startup?

Sometimes, and it depends on the state. California courts bar post-employment non-competes and customer non-solicits, but the California Supreme Court has not definitively addressed employee non-solicits, and other states enforce them. Ask every candidate for their state of employment and a self-disclosure of their invention-assignment and confidentiality agreement before the introduction, and send anything flagged to counsel. This guide is a process, not legal advice.

### Should the cofounder trial come before or after the equity conversation?

Sources disagree, so decide deliberately. One VC blog puts a trial of at least four to six weeks before any equity discussion; Antler forms teams first and trials after. Always write the starting position before the trial so the founder is not improvising. Share it before the trial if the candidate is giving up salary or serious time, and after it if the trial is light.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/guides/solo-founder-cofounder-slate-playbook*