13-Week Cash Flow Plan
it.chiribav0.1.0Updated Oct 1, 2026
13-week cash flow plan with a live Excel workbook, not just an answer in chat.
Overview
Builds a 13-week cash flow forecast from six figures and returns a live Excel workbook with formulas, the break-even week and peak funding need.
- What you can do
- Turn six inputs (opening balance, receipts, supplier payments, payroll, loan instalments, taxes) into a 13-week cash grid with closing balances.
- Identify the first week the balance turns negative, the shortfall in that week, and the peak funding need across the period.
- Convert revenue and DSO, or purchases and DPO, into weekly receipts and payments when figures are not known.
- Compare levers such as collecting 15 days earlier or moving payroll a week, with the weeks each buys.
- Produce a live Excel workbook with editable input cells and real formulas that recalculate on open.
- When to use it
- Useful when you need a quick, defensible 13-week cash forecast for a business or client and want an editable workbook rather than a chat answer. Suited to questions about when cash goes negative, how large the funding gap is, and which levers would move the date.
- Requirements
- A remote MCP endpoint at mcp.chiriba.it, or a local desktop extension file installed in Claude Desktop. No account, token or environment variables are needed for the remote server or the desktop extension. The optional Apify route requires your own Apify API token appended to the URL.
Installation
In SourceWeft
- Open 13-Week Cash Flow Plan in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.
Use with an agent
Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.
Read https://sourceweft.com/mcp/it.chiriba%2Fcashflow13/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.Other MCP clients
Add this to your client's mcpServers config.
{
"mcpServers": {
"13-week-cash-flow-plan": {
"type": "http",
"url": "https://mcp.chiriba.it/mcp"
}
}
}README
13-Week Cash Flow Forecast
Answers the question people actually open a cash flow tool to ask: which week does the cash go negative, and by how much?
Six figures in — opening cash balance, expected receipts, supplier payments, payroll, loan instalments, and VAT or taxes due. Out comes the 13-week grid, the break-even week, the peak funding need across the period, the levers that would move the date, and a live Excel workbook with real formulas you can keep working in.
Don't know your receipts? Give revenue and DSO instead and it converts. Same for purchases and DPO on the payables side. One number per line if your weeks are flat, thirteen if you have the detail.
Three ways to use it
1. Desktop extension — nothing leaves your machine
The best version, and the one to pick if the figures are a client's. The server runs locally and writes the workbook straight to your disk. No account, no token, no upload.
Download cashflow13-*.mcpb from
Releases,
then in Claude Desktop: Settings → Extensions → drag the file in.
2. Remote server — the shortest path
No account and no token. In Claude Desktop, Settings → Connectors → Add custom connector:
In Claude Code:
Listed in the official MCP registry as it.chiriba/cashflow13.
3. Apify Actor
Apify Standby requires your own Apify API token — that is the platform's rule, not ours. Append it to the URL:
If you would rather not have an account at all, use one of the two above.
What it gives back
- the week-by-week grid with the closing cash balance
- the first week the balance turns negative, and the shortfall in that week
- the peak funding need over the 13 weeks — the cumulative low point, which is a different number from the first shortfall, and the one your bank will ask for
- levers, each with the weeks it buys and the change in funding need: collecting 15 days earlier, paying suppliers 15 days later, moving payroll a week
The levers never make an outflow disappear. Moving payroll a week pushes the last payment past week 13, and the answer says so — deferred, not saved.
The workbook
Yellow cells are inputs. Everything else is a formula. Change a weekly figure and the grid, the break-even week and the funding need all move; change the start date in one cell and all thirteen weeks follow. It recalculates on open, because no cached values are written — it is the model, not a picture of one.
Weeks end on Friday by default: US practice, and where payroll and bank cut-offs fall. Configurable, and written into the workbook header so whoever receives the file knows what convention they are reading.
Your figures
No database, no history, no account, nothing kept.
- Desktop extension — the numbers never leave your computer.
- Remote server — the workbook is held in memory for 60 minutes, never written to disk, then forgotten.
- Apify — the workbook sits on Apify's storage for at most 20 minutes, then is deleted whether it was downloaded or not.
Development
Rules that shaped this
- No user-facing string is hardcoded — it goes in the localization table, in English and Italian. A test fails if a key is missing a language, or if an Italian string is a copy-paste of the English.
- Tests before code. For a financial tool, correctness is the product.
A balance of exactly zero is not negative. Money rounds to the cent at every
step, because
0.1 + 0.2does not make0.3and a plan that says "you are overdrawn" when you are not is worse than no plan. - No storage. That is also the GDPR answer.
- One engine, many tools. The next tools are new functions on this server, not new projects.
Licence
Business Source License 1.1. In plain terms:
- Use it for anything, including your own professional work — an accountant or controller running this for a client, and billing that client, is fine.
- You may not hand it to third parties for a fee so they can use it themselves — not as a product, not as a hosted service, not embedded in other software.
- Four years after each version is published, that version becomes Apache 2.0 and the restriction lapses.
Not OSI-approved, deliberately: the tool is free to use and stays free to use,
but the option to charge for it later is kept open.
For other arrangements: [email protected].
Also in this repository
Dataset Audit — you paid per
result; it tells you what you got. Duplicates, records that ignored the filters
you asked for, and the real cost per usable record, with a live Excel receipt.
Free. Source in actors/dataset-audit/, same engine as
this one.
Source: README.md at commit bd49d8f
Tools
0Version history
1- v0.1.0LatestSep 16, 2026
Score
40% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 daysRequired25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declaredRequired10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.
