Self-Hosted Two-Agent Role-Play Studio

LLMs

A web app where two AI agents each take a role and chat to complete a task you give them; runs on your own machine with Docker Compose, bound to 127.0.0.1, with your own model key. Derived from a frozen snapshot of an Apache-2.0 open-source project.

Subscribe & download See how to start

SubscriptionUS$49 per month

Last run2026-11-12

In practice

Readiness3 / 5

What you need

Few changes needed3 / 5

Who it is for

Individuals and small teams who want to try two-agent role-play conversations on their own machine and study how agents divide a task.

How to start

$ cp .env.example .env
$ docker compose up -d --build

Browse the files

Files in the product repository at the pinned commit. A few house files are left out of the zip.

File overview2,000+ paths

▸apps18

agents/
common/
data_explorer/
dilemma/

▸camel487

__init__.py
agents/
benchmarks/
bots/
caches/
configs/
data_collectors/
datagen/
datahubs/
datasets/
…

▸data4

ai_society/
code/

▸docs822

Makefile
README.md
_static/
camel.agents.rst
camel.benchmarks.rst
camel.bots.discord.rst
camel.bots.rst
camel.bots.slack.rst
camel.configs.rst
camel.data_collector.rst
…

▸examples359

__init__.py
agents/
ai_society/
benchmarks/
bots/
caches/
code/
conversion.py
custom_client_usage.py
data_collectors/
…

▸licenses2

license_template.txt
update_license.py

▸misc4

favicon.png
framework.png
logo_light.png
wechat.jpeg

▸profiling1

chatagent_overhead.py

▸services4

README.md
__init__.py
agent_mcp/

▸test264

agents/
benchmarks/
bots/
caches/
conftest.py
data_collectors/
data_samples/
datagen/
datahubs/
datasets/
…

▫ATTRIBUTION.md

▫CONTRIBUTING.md

▫Dockerfile.studio

▫LICENSE

▫Makefile

▫PRODUCT.md

▫QUICKSTART.md

▫README.ja.md

▫README.md

▫README.zh.md

▫SECURITY.md

▫SECURITY.upstream.md

▫camel_log_viewer.html

▫delivery-manifest.txt

▫docker-compose.yml

▫pyproject.toml

▫uv.lock

▫Config and hidden files 18

Latest passing run

Clone → Checkout → Toolchain → Run studio

Pre-purchase Disclosure

What you are buying

A web app where two AI agents each take a role and chat to complete a task you give them: brought up with Docker Compose on your own machine, with your own model key.

Derived from an open-source project under the Apache-2.0 licence, and not the original project itself.

Delivery is a frozen snapshot: one zip, taken from the product repository at revision 895f296.

Known limitations

The service speaks plain HTTP and is bound to the local machine only (port 8080); TLS and a reverse proxy are yours to bring for any wider access
To make the agents actually chat you need your own model provider key; the page itself opens without one
The first build needs internet access to download Python dependencies
Pinned to the delivered snapshot; it does not follow upstream.

The liability cap, the scope of what is promised and the statutory floor are governed by the Terms of Service.

Vetting record

Everything below was checked against the pinned commit 0366b64. Every verdict carries its full reasoning, an evidence link and the checker; failures are published, not hidden — do not take our word for it, verify it yourself.

2 PASS3 WARN0 FAIL

Source ReviewWARN

Read: Dockerfile.studio, docker-compose.yml, .env.example, QUICKSTART.md and the launch section of the web app entry file. The rest of the upstream Python source was not read file by file.

Script SafetyPASS

The Dockerfile and the compose file were read: the base image is pinned by digest, the container runs as a non-root user, and no download-and-pipe-to-shell pattern was seen. Other upstream scripts and examples were not read.

RecencyWARN

This is a frozen snapshot that does not follow upstream. The date of the pinned upstream commit and recent upstream activity were not checked.

LicensePASS

SPDX:Apache-2.0

The LICENSE file in the product repository is the Apache-2.0 text, and PRODUCT.md records the upstream license as Apache-2.0. The upstream copy was not compared against it.

Dependency LicensesWARN

Dependencies are declared in the Dockerfile and fetched at build time. They were not scanned for licenses.