> For the complete documentation index, see [llms.txt](https://robodex.gitbook.io/robodex-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://robodex.gitbook.io/robodex-docs/research/agent-loop-registry.md).

# Agent Loop Registry

The registry lives in:

```
packages/research/agent_loops/registry.py
```

It is the approved product-to-research bridge.

## Public functions

| Function                                                        | Purpose                        |
| --------------------------------------------------------------- | ------------------------------ |
| `register_agent_loop(name, runner)`                             | Register a loop implementation |
| `get_agent_loop(name)`                                          | Retrieve a registered runner   |
| `list_agent_loops()`                                            | Return available loop names    |
| `run_agent_loop(name, prompt, initial_state=None, config=None)` | Execute a loop by name         |

## Built-in loop loading

The registry loads built-ins lazily so import-time failures do not break the entire app. Current built-ins include:

* `creative_qd_v2`,
* `grammar_v2`,
* `grammar`.

This lazy loading is important because some research dependencies may be heavier than product startup should require.

## Runner contract

Loops follow the `AgentLoopRunner` protocol from `packages/research/agent_loops/protocol.py`:

```python
loop(prompt, initial_state=None, *, config=None) -> AgentLoopResult
```

`AgentLoopResult` contains:

* `state`: full debug or research state,
* `hitl`: backend-facing payload.

## Configuration

`AgentLoopConfig` includes:

* population,
* max attempts,
* confirmed specification,
* normalizer agent,
* rule builder agent,
* evaluator agent,
* human confirmation flag,
* extra options.

## Trace metadata

`run_agent_loop()` normalizes LangSmith or tracing metadata into returned state and HITL when available. That lets product and tests inspect trace ids without knowing how the loop was instrumented.

## Product use

Product code should use:

```python
run_agent_loop(...)
list_agent_loops()
```

It should not import `grammar_v2_loop.py` or `creative_qd_v2_loop.py` directly unless there is a narrow test or development reason.


---

# 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://robodex.gitbook.io/robodex-docs/research/agent-loop-registry.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.