> For the complete documentation index, see [llms.txt](https://bubblegum-reality.gitbook.io/bubblegum-reality-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://bubblegum-reality.gitbook.io/bubblegum-reality-docs/architecture/reality-check-ai.md).

# Reality Check (AI)

The Reality Check turns the deterministic execution snapshot into a short, bounded explanation. It never predicts prices and never recommends a trade.

## Flow

```
swap panel builds RealitySnapshot   (symbol, session, size, reference vs executable,
        │                            spread bps, cost bps, Reality Score components)
        ▼
getRealityCheck()  src/lib/ai.functions.ts — Zod-validated, identity-free
        ▼
snapshotCacheKey() → SHA-256 digest → ai_insights lookup (10 min TTL)
        │ hit  ─────────────────────────────▶ cached verdict
        │ miss
        ▼
AI gateway · google/gemini-3.6-flash · structured output
        ▼
coerceRealityCheckShape() → RealityCheckSchema → cache write → panel
```

## Contract

`RealityCheckSchema` (`src/lib/ai-schema.ts`) bounds the output: a `verdict` of `favorable | mixed | unfavorable`, a headline, a summary, factors with `label`/`detail`/`impact`/`confidence`, and cautions. Anything outside the schema is rejected as `invalid_response` rather than rendered.

`coerceRealityCheckShape()` maps known model drift (for example `name` → `label`) and fills missing `impact`/`confidence` before validation, so a small formatting wobble degrades to a valid verdict instead of an error.

## Failure mapping

| Upstream             | Client error code   |
| -------------------- | ------------------- |
| 429                  | `rate_limited`      |
| 402                  | `credits_exhausted` |
| schema/parse failure | `invalid_response`  |
| anything else        | `unavailable`       |

No snapshot contains a wallet address, account ID, or amount owned. The panel always renders `REALITY_CHECK_DISCLAIMER` alongside the verdict.


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://bubblegum-reality.gitbook.io/bubblegum-reality-docs/architecture/reality-check-ai.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.