> For the complete documentation index, see [llms.txt](https://jig.gitbook.io/sley/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://jig.gitbook.io/sley/guides/inspection.md).

# Inspection

Inspect a compiled Sley graph and structured run result without mistaking either for a full execution trace.

When a workflow takes a surprising route, you need to separate two questions: what graph did we define, and how did this particular run settle?

Sley exposes one view for each question:

* `compile().describe()` describes the graph before it runs;
* `start().result()` describes how its branches settled.

Neither is a step-by-step trace. This complete example inspects both views of one graph.

{% tabs %}
{% tab title="Python" %}

```python
import asyncio

from sley import Flow, node


@node
def greet(context):
    context.end("hello")


async def main():
    compiled = Flow(greet, name="greeting").compile()

    for element in compiled.describe()["elements"]:
        print(f"topology: {element['kind']} {element['name']}")

    result = await compiled.start({}).result()
    print(f"result: {result.status}")
    for terminal in result.terminals:
        print(f"terminal: {terminal.type}")


asyncio.run(main())
```

{% endtab %}

{% tab title="TypeScript" %}

```typescript
import { Flow, node } from '@jigging/sley'

const greet = node(
  (context) => {
    context.end('hello')
  },
  { name: 'greet' },
)

const compiled = new Flow(greet, { name: 'greeting' }).compile()

for (const element of compiled.describe().elements) {
  console.log(`topology: ${element.kind} ${element.name}`)
}

const result = await compiled.start({}).result()
console.log(`result: ${result.status}`)
for (const terminal of result.terminals) {
  console.log(`terminal: ${terminal.type}`)
}
```

{% endtab %}
{% endtabs %}

Both programs print:

```
topology: flow greeting
topology: node greet
result: completed
terminal: end
```

## Inspect the topology that will run

Compilation validates and snapshots the reachable definition. `describe()` returns detached plain data containing elements, scopes, links, entries, declared exits, concurrency, and activation limits.

The compiled Flow keeps its snapshot even if the mutable `Node` or `Flow` objects are linked differently later. Compile once when a stable definition is important; compile again when graph edits should take effect.

## Inspect how branches settled

A completed result contains final state and terminal records. A failed result also contains its controlling structured failure. Terminal sequence records settlement order, which may differ from dispatch order under concurrency.

Source activation IDs identify the activation that settled a terminal. They do not map directly to element names, so use the static description and your own domain logs when that relationship matters.

## Log domain facts where they happen

Use the host language's logger inside handlers or injected services. State, input, and terminal output move application data; logs only observe it and must not decide graph control.

Sley intentionally does not ship an event bus, observer API, or distributed trace. Add instrumentation at service boundaries, or wrap handler functions in application code when uniform timing is required.

You can now inspect topology without mistaking it for history and inspect a result without pretending it is a trace. See [Runtime semantics](/sley/reference/runtime-semantics.md) for every description and terminal field.


---

# 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://jig.gitbook.io/sley/guides/inspection.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.