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

# Mechanism graph

Curation, not discovery. A finite map of how one factor moves another.

Most "X happened because of Y" explanations in crypto are stories. LensAI keeps a hand-curated, finite map of **documented transmission channels**: the mechanical ways one factor moves another. A connection the desk states is either measured, mechanical through this graph, or explicitly labelled conjecture.

## What is in it

The graph is made of **nodes** (measured factors such as funding, depth or stablecoin float, and concepts such as the basis trade, spot buying power or risk appetite) and **directed edges** (one node acting on another, with a polarity and a written mechanism).

The current seed covers the core channels:

| Channel                               | Mechanism, in brief                                                                                                    |
| ------------------------------------- | ---------------------------------------------------------------------------------------------------------------------- |
| Funding and the basis trade           | Rich funding draws cash-and-carry capital (long spot, short perp), whose shorting presses funding back toward neutral. |
| Basis trade and ETF arbitrage         | ETF creation and redemption arbitrage connects spot ETF demand to the basis that carry traders harvest.                |
| Stablecoin float to spot buying power | Circulating stablecoins are on-chain dry powder; growth adds capacity to bid spot.                                     |
| Real yields to risk appetite          | Higher real yields lift the risk-free alternative and the discount rate, weighing on risk assets.                      |
| The dollar to risk appetite           | A stronger dollar has coincided with lower risk appetite across assets.                                                |
| Depth to gap risk                     | Thin resting liquidity near the mid price makes discontinuous moves more likely on a given flow.                       |

The graph is a seed, built to grow: liquidation cascades, auto-deleveraging and cross-venue channels come next as the data for them arrives.

## Rules

* **Curation, not discovery.** Channels are written and reviewed by people from documented market mechanics. The engine does not invent relationships.
* **Versioned and append-only.** A change is a new graph version, never an edit. Every read takes the version that was current at its as-of time, so past reads stay reproducible.
* **Permanent ids.** A node or edge id never changes meaning. A renamed concept is a new id, with the old one marked deprecated.
* **Directed only.** A two-way relationship is two edges sharing a channel.
* **Regime-aware.** An edge can state the regimes in which it applies and those in which it is known to break.
* **Strength is a prior.** Each edge carries a curated strength, labelled everywhere as a prior until the calibration record can measure it.
* **Guardrailed.** Every prose field in the graph passes the same non-advisory checks as the desk's output, in CI.

## How it is used

* **Narration:** a claim marked "mechanical" must cite a real edge id, or it is dropped.
* **Divergence:** the engine checks graph relationships between measured factors and flags those that have broken.
* **Claim Check:** a causal claim in a pasted post counts as documented only when it maps to an edge.
* **Evidence:** answers show the channels in play, with the ones they cited marked.


---

# 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.scanbylens.xyz/the-engine/mechanism-graph.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.