> For the complete documentation index, see [llms.txt](https://docs.aviator.co/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.aviator.co/verify.md).

# Verify

Aviator Verify replaces line-by-line code review with verification of *what the change is supposed to do*. You work with your agent locally. Aviator captures your intent through a single CLI call, runs every acceptance criterion against the running code, and produces a review document with verdicts, evidence, and the decisions behind the change.

The reviewer judges behavior and the calls that shaped it, starting from those rather than the diff.

You can start using Verify with code-scan alone — no preview, no infrastructure setup. A preview is what unlocks **runtime verification**: scenarios driven against your running code. Most teams get value from day one with code-scan, then add a preview when behavioral criteria start mattering.

### The flow

1. **Build with your agent.** Implement the task with Cursor, Claude, Copilot — anything you use today. No new tooling, no behavior change.
2. **Submit intent with the Aviator CLI.** When you're done, the agent runs `/verify-submit`, which captures the intent and the acceptance criteria from what was built and submits them. Set up [agent hooks](https://docs.aviator.co/verify/how-to-guides/set-up-agent-hooks.md) and it gets prompted before every PR.
3. **Record the decisions.** `/verify-submit` also drafts the calls behind the change: what was picked, what risks were accepted, and whether a person or an agent made each call. Once you confirm them, it records them on the review.
4. **Aviator verifies end to end.** Scenarios run in your preview, invariants apply automatically, code-scan handles structural checks.
5. **Review the behavior and the decisions.** The reviewer sees the intent, the verdict per criterion, the evidence, and the decisions. They approve, waive with reason, ask for another scenario, or reopen a decision.

→ [How it works](https://docs.aviator.co/verify/how-it-works.md) — the full narrative, with diagrams.

### Decisions — the calls behind the change

Verification can show that a change does what was agreed. It can't say whether what was agreed was the right call. Decisions record those calls: per account or per repo, ship the risk or fix it, do this now or later.

Each decision has a type, an outcome, a reason, and a comment thread. Humans and agents are credited separately on every entry, so a reviewer can see which positions a person took and which ones an agent took on its own.

→ [Concepts: Decisions](https://docs.aviator.co/verify/concepts/decisions.md) · [Work with decisions](https://docs.aviator.co/verify/how-to-guides/working-with-decisions.md)

### Invariants — the standing rules

Every code review has a long tail of repeated comments. *Use the structured logger. No direct user-table writes. Errors emit a metrics counter. Cap external dependencies. Stripe amounts use the Money type.* Reviewers say the same thing on PR after PR, and any one of them missed in any one review becomes a future bug or migration.

Verify turns those into **invariants** — team-defined rules in an account-level catalog that apply to every matching change automatically. When a review is created, an LLM selector picks which invariants legitimately apply to this change, and they're materialized as acceptance criteria alongside the user-supplied ones. From there they flow through the same verifier pipeline — code-scan or runtime — and produce verdicts on the same review surface.

How invariants get into your catalog:

* **Mine your PR history.** Aviator's AI reads your team's actual PR review comments and proposes invariants from the patterns it sees. The highest-leverage onboarding path — your team is already enforcing these rules in review, just not encoding anywhere.
* **Extract from your docs.** `CONTRIBUTING.md`, `LLM.md`, and similar files become drafts.
* **Adopt from templates.** Aviator's starter library covers the common categories: security baseline, observability, data access, backwards compatibility.
* **Author manually.** Anyone on the account writes the rule directly.

All AI-drafted invariants land in draft status and need [approval](https://docs.aviator.co/verify/concepts/invariants.md#owner-teams) before they go active. Failed invariant verdicts can be waived by the reviewer with a category (`false_positive`, `doesnt_apply`, `accepted_risk`, `fix_in_followup`) — every waiver is recorded in the audit trail.

→ [Concepts: Invariants](https://docs.aviator.co/verify/concepts/invariants.md)

### How review changes

A diff shows what changed. It doesn't show what the change was for, whether it works, or why it was built this way. Verify answers those separately:

* **What is it for?** The intent.
* **Does it do that?** Acceptance criteria, verdicts, and evidence.
* **Does it break our standing rules?** Invariants.
* **Why this way, and who made the call?** Decisions.

→ [How Verify reviews a change](https://docs.aviator.co/verify/concepts/review-model.md)

### Core concepts

| Concept                                                                                                                 | What it is                                                                                        |
| ----------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- |
| [Intent and acceptance criteria](https://docs.aviator.co/verify/how-to-guides/writing-effective-acceptance-criteria.md) | What the agent submits — what the change is for, and the verifiable assertions it must satisfy.   |
| [Decisions](https://docs.aviator.co/verify/concepts/decisions.md)                                                       | The calls behind a change, with the reason and who made each one, person or agent.                |
| [Invariants](https://docs.aviator.co/verify/concepts/invariants.md)                                                     | Team-defined rules applied to every matching change. Security baseline, data access, conventions. |
| [Previews](https://docs.aviator.co/verify/concepts/previews.md)                                                         | Ephemeral environments scenarios run against. Per-run, torn down after.                           |
| [Verification layers](https://docs.aviator.co/verify/concepts/verification-layers.md)                                   | How criteria, invariants, and domain contracts stack.                                             |
| [Audit and compliance](https://docs.aviator.co/verify/concepts/audit-trails-and-compliance.md)                          | Immutable trail of submissions, verdicts, and evidence, plus each decision's version history.     |

### Getting started

1. [Connect a repository](https://docs.aviator.co/verify/how-to-guides/connect-a-repository.md) — wire up GitHub.
2. [Your first verification](https://docs.aviator.co/verify/your-first-spec.md) — a 15-minute hands-on. Install the CLI, ship a small change, watch verification run. No preview needed.
3. [Setting up org invariants](https://docs.aviator.co/verify/setting-up-org-invariants.md) — codify the rules every change must respect.
4. [Creating a preview](https://docs.aviator.co/verify/how-to-guides/creating-a-preview.md) — *optional.* Unlocks runtime verification when your criteria start asking about behavior the diff alone can't answer.

### Quick links

* **Submitting:** [Aviator CLI](https://docs.aviator.co/verify/reference/cli.md) · [Set up agent hooks](https://docs.aviator.co/verify/how-to-guides/set-up-agent-hooks.md)
* **Writing for the agent:** [SKILL.md guide](https://docs.aviator.co/verify/how-to-guides/writing-a-skill-md.md) · [Effective acceptance criteria](https://docs.aviator.co/verify/how-to-guides/writing-effective-acceptance-criteria.md)
* **Configuration:** [Preview YAML](https://docs.aviator.co/verify/reference/preview-yaml.md)
* **Operations:** [Managing previews](https://docs.aviator.co/verify/how-to-guides/managing-previews.md) · [Seed data for previews](https://docs.aviator.co/verify/how-to-guides/seed-data-for-previews.md)
* **Reviewing:** [Review verification on the pull request](https://docs.aviator.co/verify/how-to-guides/verify-on-github.md) · [Work with decisions](https://docs.aviator.co/verify/how-to-guides/working-with-decisions.md)
* **Recording decisions:** [Record decisions with /aviator:decisions](https://docs.aviator.co/verify/how-to-guides/recording-decisions.md)
* **Reading results and fixing failures:** [Understanding verification results](https://docs.aviator.co/verify/how-to-guides/understanding-verification-results.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.aviator.co/verify.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
