> For the complete documentation index, see [llms.txt](https://zkwasmdoc.gitbook.io/delphinus-zkwasm/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://zkwasmdoc.gitbook.io/delphinus-zkwasm/c1_beginner/c1_quickstart/c5_aggregate.md).

# Aggregate (Batch) proofs

Suppose that we have generated a guest proof with its loading info and stored in the following directory.

```
.
├── Makefile
├── output.wasm
├── params
  ├── K22.params
  ├── testwasm.circuit.finalized.data
  └── testwasm.zkwasm.config
├── output
  ├── traces
  ├── testwasm.loadinfo.json
  ├── testwasm.0.transcript.data
  └── testwasm.0.instance.data
```

Also suppose the overall information of the proof is stored in the zkwasm.loadinfo.json which looks like the following

```
{
  "k": 18,
  "proofs": [
    {
      "circuit": "testwasm.circuit.finalized.data",
      "instance_size": 4,
      "witness": "testwasm.0.witness.json",
      "instance": "testwasm.0.instance.data",
      "transcript": "testwasm.0.transcript.data"
    }
  ],
  "param": "K18.params",
  "name": "testwasm",
  "hashtype": "Poseidon"
}
```

Although the guest circuit proof is constant size, it is still big and its verify algorithm is a bit costy. In ZKWASM we have a structured layer of batching multi proofs into a simple small proof to reduce the cost of onchain verification. (Please refer to [proof batch](/delphinus-zkwasm/c2_advance/c3_proofgen/c2_batch.md))

Here, for simplification, we use ZKWASM batcher to batch the above proof into a small one.

Suppose that your project directory is the following:

```
.
├── Makefile
├── output.wasm
├── params
  ├── K22.params
  ├── testwasm.circuit.finalized.data
  └── testwasm.zkwasm.config
├── output
  ├── traces
  ├── testwasm.loadinfo.json
  ├── testwasm.0.transcript.data
  └── testwasm.0.instance.data
```

1. write a simple empty batch config `batchconfig.json`:

```
{
    "equivalents": [
    ],
    "expose": [
    ],
    "absorb": []
}
```

2. clone the repo <https://github.com/DelphinusLab/continuation-batcher>
3. run the continuation batcher in batch mode:

```
cargo run --release --features cuda -- --params ./params --output ./output batch -k 22 --challenge sha --info output/testwasm.loadinfo.json  --name proofbatch --commits batchconfig.json
```

This will generates a batch proof `proofbatch.loadinfo.json` and the file structure in your project becomse the following

```
├── Makefile
├── output.wasm
├── params
  ├── K22.params
  ├── proofbatch.circuit.data
  ├── proofbatch.vkey.data
  ├── testwasm.circuit.finalized.data
  └── testwasm.zkwasm.config
├── output
  ├── proofbatch.0.aux.data
  ├── proofbatch.0.instance.data
  ├── proofbatch.0.transcript.data
  ├── proofbatch.loadinfo.json
  ├── testwasm.loadinfo.json
  ├── testwasm.0.transcript.data
  └── testwasm.0.instance.data
```

while `the proofbatch.circuit.data` is the verify circuit of the batcher and `proofbatch.loadinfo.data` describes the proof of the batching result. Since design a product ready rollup process is much more complicated than just batching a few proofs of a single circuit, please refer to [advanced topics of building a rollup application](/delphinus-zkwasm/c2_advance.md) for more details.


---

# 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://zkwasmdoc.gitbook.io/delphinus-zkwasm/c1_beginner/c1_quickstart/c5_aggregate.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.