# Amortization Schedule

> Row-by-row loan amortization schedules: monthly or daily accrual, extra principal, PMI, biweekly.

- name: com.gumballtools/amortize
- version: 0.1.0
- connection class: R0 (autonomous)
- trust: 89/100 flags: duplicate-repo
- quality: 86/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/com.gumballtools%2Famortize
- tools: https://api.protogrid.dev/v1/servers/com.gumballtools%2Famortize/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "amortize": {
      "type": "http",
      "url": "https://amortize.gumballtools.com/api/mcp"
    }
  }
}
```


## Trust

- hygiene: 80
- liveness: 100
- freshness: 85
- provenance: 85
- drivers: +repository +dns-namespace +website +reachable +uptime-100% ~updated-45d-ago -duplicate-repo(16 names)
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 86/100 (good)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 79 (weight 30)

- pass `tools.descriptions`: every tool has a description
- warn `tools.description_length`: 1 tools have descriptions over 1,024 characters, which crowds the context
- pass `tools.schemas`: every tool has an object input schema
- pass `tools.annotations`: 1 of 1 tools declare readOnlyHint or destructiveHint
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: amortize (destructiveHint, idempotentHint)
- pass `tools.token_cost`: about 1,361 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: no tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 1, about 1,361 tokens to load them all
- tool set hash: be67220883527d6609469746d1dc3d50
- badge: [![protogrid quality](https://protogrid.dev/badge/com.gumballtools/amortize.svg)](https://protogrid.dev/servers/com.gumballtools/amortize)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

_No tool definition changes recorded._

## Remotes

- https://amortize.gumballtools.com/api/mcp (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (1)

- `amortize`: Builds the full row-by-row amortization schedule for a loan — interest, principal, PMI, and running balance for every payment — instead of just the monthly payment amount.

Use this whenever a row-by-row schedule matters: "what is my balance after payment 47", "what does an extra $200/month do to my payoff date", "when does PMI drop off", or "give me the whole amortization table". Do not do this arithmetic by hand or in your head. Two things go wrong reliably when a model tries:

1. THE LAST ROW. The textbook payment formula only amortizes to exactly zero in infinite precision. Rounded to cents and carried over up to a few hundred rows, a hand-rolled schedule drifts and ends with a balance of a few cents to a few dollars instead of zero — exactly the row a person checks against their real statement. This tool recomputes the final row as "remaining balance + that row's interest" so it always ends at zero, and reports the delta.
2. COMPOUNDING CONVENTION. Fixed mortgages compound monthly (rate/12). HELOCs, most auto loans, and most private student loans accrue interest DAILY on the actual number of days between payments, which is a genuinely different schedule, not a rounding variant. Guessing the wrong one silently produces a plausible but wrong table.

Input: `principal`, `annualRatePercent` (e.g. 6.5), `termMonths` (integer), and `compounding` ("monthly" or "daily") are required. For `compounding: "daily"`, supply EITHER `startDate` (YYYY-MM-DD — the tool derives real day counts and leap years from the calendar itself) OR `dayCountBasis` ("30/360" | "actual/365" | "actual/actual") for a hypothetical loan with no dates — never both are needed, and supplying neither is refused rather than guessed. Optional: `paymentFrequency` ("monthly" default | "biweekly", which requires daily compounding — pass `paymentOverride` if the actual billed payment differs from the textbook one, e.g. a HELOC's quoted minimum; `extraPrincipal` as `{type:"oneOff",amount,atPaymentNumber}` or `{type:"recurring",amount,startingPaymentNumber?}`; `pmi` as `{monthlyAmount,originalValue,cancelAtPercent?}` (78 or 80, default 78) to model PMI dropping off.

Refuses rather than guesses: a payment that does not exceed the first period's interest (negative amortization) is a named error with the minimum amortizing payment attached, not a schedule that never ends. An unparseable or impossible date is refused, not reinterpreted.

Returns the full schedule, total interest and PMI paid, the payoff date or period, the final payment amount and how it differs from the standard one, and — when extra principal was given — the same schedule without it, so the periods and interest saved are explicit. A `disclaimer` field travels with every response: this is a planning estimate, not a lender payoff quote, and PMI cancellation here uses your actual balance, which a servicer is legally permitted to compute off the original schedule instead. Relay the warnings, not just the numbers.