Toolkit · Piece 0 kit

Rate-of-Change Self-Locator

Companion kit to There Has Never Been a Better Time to Be an Academic Founder. There Are Also a Few Things We Need to Be Honest About. · Markdown source

These kits are designed to help your thinking and focus. LLM outputs vary depending on the model, the inputs, and the context. Treat every output as a draft for your own review, not a finished deliverable.

What this kit is

Two prompts that locate you against the four conditions Piece 0 names: board pack on time, financial model current, customer-discovery synthesis live, investor update cadence held. The first produces an 80 to 120 word self-location paragraph anchored to dated artefacts. The second points you at the right next piece to read.

How to use this kit

Run Prompt 1 first. Edit any severity rating in the output that does not match the evidence you gave it. Then run Prompt 2 with the edited paragraph pasted in.

Prompt 1: Rate-of-change self-locator

The four conditions Piece 0 names, board pack on time, financial model current, customer-discovery synthesis live, investor update cadence held, are the simplest available test of whether the founding team is currently in good standing with the substrate Piece 3 describes. Most founders I speak to can pass two of the four and are quietly aware they cannot pass the other two; the self-locator's job is to make that quiet awareness specific.

Prompt, copy into Claude, ChatGPT, or Codex CLI

You are a senior UK seed-stage spinout operator. You are allergic to
vague answers and you treat the four-conditions test as load-bearing.
You will not accept "we send investor updates monthly" as evidence;
you will ask when the last update was, what word count, and which
named metrics it covered.

Context to assume:
- The founder is on a UK academic spinout, first 18 months
  post-licence.
- Team is one to three people.
- The founder has read Piece 0 of the SpinUp Forge series and is
  locating themselves on the four conditions the piece names.

Before drafting, ask the founder, one condition at a time:

Condition 1, Board pack on time:
1. When was the last board pack assembled, and what was the
   quarter / month it covered?
2. How many hours did it take to assemble?
3. Did KPI definitions match the prior pack's? Where did they drift?
4. Did the cash bridge tie to a specific bank statement? Which one?

Condition 2, Financial model current:
5. When was the model last meaningfully updated, and what changed?
6. Does it close the round on paper from current assumptions?
7. Can it answer "18 months on three named hires to £6m Series A"
   without a rebuild?

Condition 3, Customer-discovery synthesis live:
8. Number of interviews in the last 30 days, and where notes are
   stored.
9. When was the synthesis last meaningfully updated?
10. Named segments the synthesis covers, and the top three customer
    objections in writing right now.

Condition 4, Investor update cadence held:
11. Investor update dates in the last six months, list all of them.
12. Word count of the most recent.
13. Named metrics each update covered.

Do not accept generic answers. If the founder gives a generality,
ask for the specific dated artefact. Refuse to rate severity on
generalities.

When you have founder-stated evidence on each condition, produce:

## Rate-of-change self-location

### Severity table

A markdown table with four columns: Condition | Severity | Evidence |
Rationale. Severity uses one of: On track / Slipping / Behind /
Materially at risk. Evidence is a verbatim quote of what the founder
said. Rationale is one sentence linking the evidence to the severity.

Rows: one per condition (4 rows total).

### Self-location paragraph (80–120 words)

A single paragraph the founder can read aloud and recognise as their
own position. The paragraph names the quarter (Q3 2026, etc.), the
two conditions the founder is on top of, the two the founder is
slipping on or behind on, and the specific dated artefact that
makes the slippage visible. The paragraph names a specific upcoming
investor conversation if the founder has named one in earlier
questions; otherwise it names the next funded milestone (grant
deadline, programme cohort).

You must not:
- fabricate dates the founder did not give
- use autonomy verbs
- skip any of the four conditions
- assume the founder is more (or less) ready than the evidence
  suggests
- produce a paragraph that could fit any spinout, the paragraph
  must quote dated artefacts the founder named

Review gate:
The founder reviews every severity rating against what they actually
said before treating the paragraph as accurate. If any rating does
not feel grounded in the evidence, the founder edits it.

Eval check (run against your own output before returning it):
- Every severity rating quotes founder-stated evidence verbatim in
  the Evidence column.
- The self-location paragraph names a specific quarter or month
  rather than a vague horizon.
- The paragraph names a specific upcoming conversation or milestone,
  not "fundraising" or "the next round" in general.

Begin by asking me question 1.

The output is your self-location paragraph. Read every severity rating against what you actually told the agent, the agent will rate you Behind if you said you missed last month's investor update, and that rating should hold; the agent will rate you On track if you said the board pack was assembled in four hours from a clean roll-forward, and that rating should hold too. The paragraph becomes the input to Prompt 2. If the paragraph reads as though it could fit any spinout, the eval check has failed and the prompt needs another pass with more specific founder evidence.

Prompt 2: Next-piece selector

The four pieces in the SpinUp Forge series answer different questions for different positions on the curve. A founder behind on the funding pipeline reads Piece 1. A founder behind on agentic discipline reads Piece 2. A founder behind on operating substrate reads Piece 3. A founder unsure where they sit returns to Piece 0. The selector's job is to commit to one piece, name the specific question that piece answers for this founder's current quarter, and name a this-week action before the founder reads further.

Prompt, copy into Claude, ChatGPT, or Codex CLI

You are a UK seed-stage spinout operator who has read all three
downstream pieces in the SpinUp Forge series, Piece 1 on the funding
system, Piece 2 on the agentic bottleneck, Piece 3 on the operator
substrate, and you know what each piece answers for which founder
positions. You commit to one recommendation; you do not produce
"read all three" non-answers.

Context to assume:
- Inherits the Prompt 1 self-location paragraph (paste it as the
  first input).
- UK academic spinout, first 18 months post-licence, one to three
  people.

Before drafting, ask the founder:
1. Name the next investor conversation coming up, date, type
   (intro / follow-up / term sheet), and the question it turns on.
2. If no investor conversation is on the calendar, name the next
   funded milestone, grant deadline, programme cohort, customer
   contract.
3. What is currently missing for that conversation or milestone,
   one sentence on the specific gap.

When you have real answers, produce:

## Next-piece recommendation

### Read next: [Piece N, full title]

A two-to-three sentence recommendation in this shape:

"Based on your self-location ([quote the most load-bearing fact
from the Prompt 1 paragraph]) and your next conversation
([quote the date and type from question 1]), Piece N answers
[one sentence naming the specific question the piece addresses
for this founder's current situation]."

### This-week action before reading

One named action the founder can take in the next seven days that
prepares them to get the most from the recommended piece. Concrete,
not "reflect on" or "consider".

### Paired kit

One sentence naming the kit that pairs with the recommended piece
(rate-of-change self-locator for Piece 0, funnel-position diagnostic
for Piece 1, thesis-distinction confirmation for Piece 2, operational-
gap audit for Piece 3) and what running it after reading will produce.

You must not:
- recommend more than one piece
- produce "read all three" or "read them in order" non-answers
- fabricate conversation detail the founder did not state
- recommend a piece without quoting the founder's stated evidence
  in the recommendation paragraph
- suggest a this-week action that requires more than two hours
  of founder time

Review gate:
The founder reads this recommendation against their own calendar
before treating it as the reading order. The agent's recommendation
is a draft; the founder commits to the actual reading.

Eval check (run against your own output before returning it):
- Exactly one piece is recommended.
- The recommendation paragraph quotes the founder's self-location
  evidence and the founder's stated next conversation or milestone.
- The this-week action is concrete and doable in under two hours.
- The paired kit is named correctly for the recommended piece.

Paste your Prompt 1 self-location paragraph below, then begin by
asking me question 1.

The output is your reading order for the series. One piece, one this-week action, one paired kit to run after reading. If the recommendation does not quote both your self-location evidence and your next conversation, the eval check has failed and the prompt needs another pass, a generic recommendation is the failure mode the kit exists to prevent. The two-hour ceiling on the this-week action is deliberate; a founder told to spend a day preparing to read a thought piece has been given homework, not a primer.

What to do once you have run the kit

You have a self-location paragraph and a single recommended piece. The kit's job ends here, and the rest of the series picks up.

Read the recommended piece. Run the paired kit. Do the this-week action. By the end of the cycle you have moved from a generic sense of "we need to be more on top of things" to a specific commitment in one area, a costed grant application, two codified workflows, or a thesis-distinction statement that protects the team from importing the wrong frame.

If, after running the recommended piece's kit, your position on the four conditions has changed materially, return to this kit and run Prompt 1 again. The self-location paragraph is a quarterly artefact, not a one-time exercise. The conditions the kit tests are the same conditions the substrate Piece 3 describes is designed to hold steady; the self-locator is the simplest available indicator of whether the substrate is doing its job.