# Foundry

> Foundry finds out how software actually behaves — mapping undocumented systems, reconstructing APIs, generating SDKs and verifying every finding with tests — so teams can change, integrate, modernize and connect AI agents with confidence.

Foundry shows what your software actually does — and proves it. Foundry is a product of Maxint (Maxint Inc., https://maxint.com), not a separate company. Contact: contact@maxint.com.

## What Foundry does

- **Map undocumented systems**: Find out how software nobody fully documented actually behaves: its interfaces, dependencies and hidden rules. Delivered as: Behavior report · dependency map.
- **Reconstruct APIs**: Rebuild an accurate interface specification from how a system really responds, not from what the docs claim. Delivered as: OpenAPI specification.
- **Generate SDKs and integration clients**: Typed client code for systems you need to integrate with, shipped with tests that prove it matches. Delivered as: SDK · conformance tests.
- **Verify behavior**: Repeatable tests confirm each finding. Anything not tested is labeled as observed, inferred or open. Delivered as: Test suite · verification results.
- **Assess change impact**: Know what a change, migration or replacement will affect, and which behavior has to survive it. Delivered as: Change-impact assessment.
- **Make systems agent-ready**: Map what a system lets an AI agent do, and give agents tested interfaces with limits you set. Delivered as: Capability map · tested interfaces.
- **Watch for change**: Re-analyze when a system or dependency changes, flag what moved and re-test what you rely on. Delivered as: Change reports · compatibility checks.

## Evidence states

Every Foundry finding carries one of four states, so nothing is presented as more certain than it is. Analysis is not a security certification and does not establish authorization. See [Approach](https://foundry.maxint.com/approach).

- **Verified**: Confirmed by a repeatable test against the system. Build on it. The test ships with the deliverable so you can run it again.
- **Observed**: Seen directly in the running system, not yet reproduced by a test. Rely on it with care. Ask for a test where a decision depends on it.
- **Inferred**: Reasoned from code, data or indirect evidence. Not seen, not tested. Treat it as a hypothesis. Confirm before consequential use.
- **Open**: We looked and could not establish it. Treat it as risk. Decide whether closing it is worth the effort.

## How Foundry is delivered

[Foundry Intelligence](https://foundry.maxint.com/intelligence) is the only way to buy Foundry. It is a service: Maxint operates the Foundry engine and delivers the results. The engine is not licensed to customers, so there is nothing to install or operate.

- Defined investigations, verified artifacts (behavior reports, interface specifications, SDKs, dependency maps, conformance tests) and ongoing intelligence as systems change.
- Every engagement has a defined scope, target count, deliverables and acceptance criteria.
- When work must stay inside infrastructure the customer controls, it can be scoped as a Controlled Environment Engagement. Feasibility and data flows are confirmed during scoping, before Maxint commits to it; no blanket data-residency guarantee is made.

## Pricing

Foundry is sold as a service: Maxint operates the engine and delivers the results. An Intelligence Assessment is $7,500; Verified Intelligence engagements are $25,000–$40,000; Continuous Intelligence is $3,000–$8,000 a month ($36,000–$96,000 a year); an Enterprise Intelligence Program is $75,000–$200,000+ a year; Strategic Intelligence projects are $50,000–$150,000+. Controlled-environment work is a fixed fee or about 25–75% above the equivalent standard engagement. Prices in USD. Details: [Pricing](https://foundry.maxint.com/pricing).

## Use cases

- [Understand](https://foundry.maxint.com/use-cases/understand-software): map how an undocumented or inherited system really behaves.
- [Protect](https://foundry.maxint.com/use-cases/protect-software): know what a change will affect and which behavior must survive it.
- [Advance](https://foundry.maxint.com/use-cases/advance-software): tested interfaces for integrations, modernization and AI agents.

## Who it is for

- [Engineering leaders](https://foundry.maxint.com/for/engineering-leaders): Make critical engineering decisions with fewer unknowns.
- [Engineers](https://foundry.maxint.com/for/engineers): Understand the behavior before you rebuild the interface.
- [Security and architecture](https://foundry.maxint.com/for/security-architecture): Know what a system exposes before extending what can access it.
- [AI agent builders](https://foundry.maxint.com/for/ai-agent-builders): Build agent workflows on interfaces you can understand and test.
- [Enterprise and regulated teams](https://foundry.maxint.com/for/enterprise): Software intelligence under your controls.

## Resources

Each guide is also available as Markdown at the same URL with `.md` appended.

- [Discovery, verification and security assurance are three different things](https://foundry.maxint.com/resources/discovery-verification-and-security-assurance): What behavioral discovery establishes, what verification adds, and why neither is a security audit or a compliance result.
- [How to build an inventory of undocumented dependencies](https://foundry.maxint.com/resources/inventory-of-undocumented-dependencies): A method for finding what a system calls, reads, writes and waits on, when nobody wrote it down, and for recording how sure you are about each item.
- [How to define an effective software intelligence engagement](https://foundry.maxint.com/resources/defining-a-software-intelligence-engagement): What to pin down before anyone analyzes a system for you: the decision, the authorized target, data handling, success criteria, deliverables and limits.
- [How to investigate an undocumented system without confusing inference with fact](https://foundry.maxint.com/resources/investigating-undocumented-systems): A working method for learning what an undocumented system does, and for keeping what you saw separate from what you assumed.
- [How to measure the cost of repeated engineering investigation](https://foundry.maxint.com/resources/measuring-the-cost-of-repeated-investigation): A method for counting the time engineers spend re-learning how a system works, so the cost of undocumented behavior can be compared with the cost of fixing it.
- [How to preserve business behavior during modernization](https://foundry.maxint.com/resources/preserving-business-behavior-during-modernization): How to find the behavior a legacy system's users depend on, decide what must survive a rewrite or migration, and prove afterwards that it did.
- [What a useful software behavior report should contain](https://foundry.maxint.com/resources/what-a-behavior-report-should-contain): The sections, evidence labels and level of detail that make a behavior report usable by engineers, not just readable by managers.
- [What to establish before connecting an AI agent to a sensitive system](https://foundry.maxint.com/resources/before-connecting-an-agent-to-a-sensitive-system): The facts, boundaries and tests to have in place before an AI agent is allowed to act on a system that matters.
- [Why generated SDKs need conformance tests](https://foundry.maxint.com/resources/why-generated-sdks-need-conformance-tests): Why an SDK generated from a specification can still be wrong, and what a conformance suite must cover before anyone builds on it.
- [Wrap, replace, or modernize? Deciding what to do with a legacy interface](https://foundry.maxint.com/resources/wrap-replace-or-modernize): A decision framework for legacy interfaces: when to put a facade in front, when to rebuild, and when to change the interface itself.

## Questions

### What is Foundry?

Foundry is a software intelligence product from Maxint. It finds out how software actually behaves, verifies what it finds with tests, and turns that into specifications, dependency maps, SDKs and test suites your team can use.

### Who is Foundry for?

Engineering leaders and engineers responsible for systems they did not build or cannot fully see into; security and architecture teams deciding what to connect; teams building AI agents and integrations; and organizations planning a modernization.

### How is Foundry different from a code scanner or an AI coding assistant?

Scanners look for known problems. Coding assistants write code from what you tell them. Foundry establishes what an existing system actually does, and tells you how strongly each finding is supported: verified by test, observed, inferred, or still open.

### What do I get from an engagement?

A defined package agreed during scoping. Typically a behavior report, interface specifications such as OpenAPI, a dependency map, conformance tests, known limitations and recommended next steps.

### Is a Foundry analysis a security audit?

No. Foundry shows how a system behaves and how well each finding is supported. That informs security decisions, but it is not a security certification and does not make a system secure.

### Do we have to install or operate anything?

No. Foundry is a service: Maxint operates the engine and delivers the findings and artifacts. When work has to happen inside infrastructure you control, it can be scoped as a Controlled Environment Engagement, once feasibility and data flows are confirmed.

### What happens to our code and data?

Before any analysis, we agree in writing what we receive, where it is processed, how long it is kept and when it is deleted. We only analyze systems you are authorized to have analyzed.

### How much does it cost?

An Intelligence Assessment is $7,500. Verified Intelligence engagements are $25,000–$40,000, Continuous Intelligence is $3,000–$8,000 a month, and Enterprise Intelligence Programs are $75,000–$200,000+ a year. Full pricing is at /pricing.

## Optional

- [Full text for language models](https://foundry.maxint.com/llms-full.txt): everything above plus every guide, persona page and price in one file.
- [Approach](https://foundry.maxint.com/approach): method, evidence states and limits.
- [Contact](https://foundry.maxint.com/contact)