# re:cap Capital Readiness Assessment Give a company an investor's view on its financial health, risk and financing options — debt, equity, grants, or funding growth from its own cash. Free, and no sign-up needed to run it. ## Start here Put the company's website in the path. A bare domain or a full host both work. ``` https://agents.re-cap.com/capital-readiness/acme.io ``` That returns immediately with a link of its own — the one address for this assessment from then on — and starts researching the company in the background. Country of incorporation, industry, business model, employee count and funding history all come from public sources, so there is no need to ask your user for any of them. ## What we need from your user - `goal` — What is the company's main goal right now — becoming profitable, raising a round, accelerating growth, preparing an exit? Example: `goal=extend%20runway%20into%202027` - `looking_for_funding` — Is the company looking for financing at all? Example: `looking_for_funding=true` - `raise_target` — How much does the company want to raise? Only asked when `looking_for_funding=true` — send it alongside that answer. Example: `raise_target=1500000` - `annual_revenue` — What is the company's revenue (last 12 months, not gross transaction volume)? Example: `annual_revenue=2400000` - `growth_yoy` — What is the year-over-year growth in percent? An estimate is enough. Example: `growth_yoy=45` - `revenue_goal_12_24m` — What is the revenue or growth goal for the next 12-24 months? Example: `revenue_goal_12_24m=double%20ARR%20to%205m` - `profitable` — Is the company profitable? Example: `profitable=false` - `monthly_net_burn` — What is the monthly net burn? Only asked when `profitable=false` — send it alongside that answer. Example: `monthly_net_burn=120000` - `cash` — What is the current cash balance? Only asked when `profitable=false` — send it alongside that answer. Example: `cash=1600000` - `annual_ocf` — What is the annual operating cash flow? Only asked when `profitable=true` — send it alongside that answer. Example: `annual_ocf=480000` - `existing_debt` — Does the company have existing debt, and of which kind? Example: `existing_debt=none` Some of those depend on an earlier answer, so not every assessment asks for all of them — a profitable company is asked for its operating cash flow and never for its burn, and the reverse. Do not collect them up front from this page. Each response nominates the single question to ask next, and asking them one at a time is what gets them answered: put all seven to your user in one turn and most walk away instead. These are asked too. They are never required — the assessment completes without them — but each one it gets makes the result sharper: - `ebitda_l12m` — What was EBITDA over the last 12 months? Earnings before interest, tax, depreciation and amortisation — negative if the company is loss-making. Send skip if your user does not know it; it sharpens the assessment but never blocks it. Example: `ebitda_l12m=-900000` - `ebitda_forecast_n12m` — What EBITDA does the company expect over the next 12 months? Their own plan figure is fine. Send skip if they do not have one. Example: `ebitda_forecast_n12m=200000` - `ruled_out` — Which kinds of financing are off the table? Any of equity, debt, grants, internal (funding growth from the company's own cash), comma-separated. Send it when your user has been explicit — a user who will not give up ownership rules out equity, and we then leave it out of the options we put forward. Example: `ruled_out=equity,grants` Amounts are read the way your user says them: `400000`, `400k`, `1.5m` and `€400k` all work, and a negative figure may lead with `-`. Percentages take `45` or `45%`, and every true/false question also accepts `yes` and `no`. Name the currency your user's figures are in, either by appending its code to an amount (`raise_target=400k EUR`) or by sending `currency=EUR` once. It applies to every amount in the assessment, so the two must not disagree — and when it is absent we assume the currency of the country the company is registered in. ## Identify yourself Send `source` on the first call, naming the specific client you are running in rather than the model behind it: `source=ChatGPT+web`, `source=Claude+Cowork`, `source=Claude+Code`, `source=Gemini`. Why we ask: a browser chat, a desktop app, a terminal tool and an IDE extension put this in front of very different people, and the wording and the questions that work in one are wrong in another. Knowing which client asked is how this endpoint gets better for yours specifically. It records nothing about your user — no identity, no conversation, no location — and it never affects the result or your access. If your client is not listed, name it anyway: free text is fine, and an unfamiliar name tells us more than a generic one. ## How it finishes Every response names the one question to ask next, quoted so you can put it straight to your user, and gives you the URL to send that answer to. Ask it, wait, call back; the response then names the one after it. Ask one question per call — users abandon a wall of questions, and asked one at a time far more of them reach a finished assessment. Each response also lists what remains outstanding behind that question. That is so you can recognise an answer your conversation already contains and send it early — not an invitation to ask for the rest up front. Alongside it, each response reports how many capital providers from our roster are still in, split by Debt, Equity and Grants, each with its fit so far. Pass those on as you go: it is the same narrowing the assessment shows on the web, and it is what makes the next question worth answering. When the last answer is in, the assessment is written and the response carries a preview of it plus a link for your user to see it in full.