> ## Documentation Index > Fetch the complete documentation index at: https://docs.rllm-project.com/llms.txt > Use this file to discover all available pages before exploring further. # Cookbooks > End-to-end AgentFlow examples that ship as installable plugins A **cookbook** is a small Python package that ships an `AgentFlow` + an `Evaluator` together with a `prepare_data.py` and a pair of `train_{tinker,verl}.sh` scripts. Each cookbook is self-contained: install it once with `pip install -e cookbooks/`, and the rLLM CLI discovers the agent and the evaluator by name through Python entry points. ```bash theme={null} rllm eval --agent --evaluator rllm train --agent --evaluator ``` Each cookbook is a plain async function that calls an OpenAI-compatible endpoint and returns an `Episode`. See [AgentFlow & Evaluator](/core-concepts/agentflow-evaluator) for the protocol. ## Available cookbooks | Cookbook | Pattern | Description | | --------------------------------------------------- | ------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | [`frozenlake`](/cookbooks/frozenlake) | Multi-turn gym | Navigates procedurally-generated FrozenLake grids; demonstrates direct integration with a Gymnasium environment. | | [`math`](/cookbooks/math) | Single-turn | Solves competition / textbook math with `\boxed{}` answer extraction. | | [`math_tool_agent`](/cookbooks/math_tool_agent) | Multi-turn + tool | Solves math with a calculator tool via native OpenAI function calling. | | [`agent_frameworks`](/cookbooks/agent_frameworks) | Multi-turn + tool (4 frameworks) | Same math + calculator task built four ways — LangGraph, OpenAI Agents SDK, smolagents, Strands — each in a single AgentFlow file. Use as a template for plugging your own framework into rLLM. | | [`deepcoder`](/cookbooks/deepcoder) | Single-turn | Competition coding with `\`\`\`python\`\`\`\` answer extraction; the evaluator runs the code against hidden tests. | | [`finqa`](/cookbooks/finqa) | Multi-turn + 4 tools | Financial QA over SEC 10-K tables: `get_table_names` / `get_table_info` / `sql_query` / `calculator`, judge-LLM grading. | | [`geo3k`](/cookbooks/geo3k) | VLM single-turn | Multimodal geometry on the Geometry3K dataset. | | [`solver_judge_flow`](/cookbooks/solver_judge_flow) | Multi-agent | Solver–judge pattern on the countdown task. | | [`terminal_bench`](/cookbooks/terminal_bench) | CLI-only, sandboxed harness | Evaluates Harbor's Terminus-2 terminal agent on all 89 Terminal-Bench 2.0 tasks via Modal sandboxes, snapshots, and pass\@k. No plugin to install — the whole run is `rllm` CLI commands. | | [`migrationbench`](/cookbooks/migrationbench) | Remote AgentCore runtime, sandboxed harness | Trains a Java 8→17 migration agent that runs inside an AWS Bedrock AgentCore Runtime. The agent ships from [agentcore-rl-toolkit](https://github.com/awslabs/agentcore-rl-toolkit); rLLM drives rollouts over the network. | Click into any row for the deep-dive page — install flow, dataset, eval and train commands, and key code snippets. The full source, including the launch scripts, lives at [`cookbooks/`](https://github.com/rllm-org/rllm/tree/main/cookbooks) on GitHub. ## Anatomy of a cookbook Every cookbook follows the same shape: ``` cookbooks// ├── README.md # walkthrough ├── pyproject.toml # entry-point declarations ├── _flow.py # the AgentFlow function ├── _eval.py # the Evaluator function (or evaluator.py) ├── prepare_data.py # DatasetRegistry registration / HF download ├── train.py # Hydra entry point used by train_*.sh ├── train_tinker.sh # single-machine LoRA training ├── train_verl.sh # distributed multi-GPU training └── test.py # unit tests ``` ### Entry-point declaration In `pyproject.toml`, the cookbook registers its flow and evaluator under two well-known groups: ```toml theme={null} [project.entry-points."rllm.agents"] my_agent = "my_flow:my_flow" [project.entry-points."rllm.evaluators"] my_eval = "my_eval:my_evaluator" ``` The CLI's `--agent ` and `--evaluator ` flags resolve through these groups (see [`rllm.eval.agent_loader`](https://github.com/rllm-org/rllm/blob/main/rllm/eval/agent_loader.py) and [`rllm.eval.evaluator_loader`](https://github.com/rllm-org/rllm/blob/main/rllm/eval/evaluator_loader.py)). ### Module-name collision gotcha Top-level Python module names must be unique across **all** installed cookbooks — `pip install -e` puts each cookbook's modules at the import root. If two cookbooks both ship a top-level `evaluator.py`, only one wins. Convention: prefix module names with the cookbook name. `cookbooks/math/` ships `math_flow.py` + `math_eval.py`, not `flow.py` + `evaluator.py`. `cookbooks/finqa/` ships `finqa_flow.py` + `finqa_eval.py` + `finqa_tools.py` + `finqa_constants.py`. ## Install + run ```bash theme={null} # 1. Install rLLM with the backend you want to train on uv pip install -e ".[tinker]" # or .[verl] # 2. Install the cookbook (--no-deps reuses rllm's pinned deps) uv pip install --no-deps -e cookbooks/math # 3. Verify discovery rllm agent list # should list "math" # 4. Pull the dataset and run rllm dataset pull math500 rllm eval math500 --agent math --evaluator math --max-examples 20 ``` The same flow works for any cookbook — substitute the cookbook name and dataset. ## Authoring a new cookbook The cleanest starting point is to copy an existing cookbook that matches your interaction shape: | Your agent looks like… | Copy from | | ---------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------- | | One LLM call, parse answer | [`cookbooks/math`](https://github.com/rllm-org/rllm/tree/main/cookbooks/math) | | Multi-turn with a tool | [`cookbooks/math_tool_agent`](https://github.com/rllm-org/rllm/tree/main/cookbooks/math_tool_agent) | | Multi-turn built on a third-party framework (LangGraph, Agents SDK, smolagents, Strands) | [`cookbooks/agent_frameworks`](https://github.com/rllm-org/rllm/tree/main/cookbooks/agent_frameworks) | | Multi-turn with multiple tools + complex grading | [`cookbooks/finqa`](https://github.com/rllm-org/rllm/tree/main/cookbooks/finqa) | | Drives a Gym environment | [`cookbooks/frozenlake`](https://github.com/rllm-org/rllm/tree/main/cookbooks/frozenlake) | | Solver + judge multi-agent | [`cookbooks/solver_judge_flow`](https://github.com/rllm-org/rllm/tree/main/cookbooks/solver_judge_flow) | Then: 1. Rename the modules (prefix with your cookbook name to avoid collisions). 2. Rewrite the flow body — call the LLM, drive your loop, return an `Episode` with the model's final answer in `episode.artifacts["answer"]`. 3. Rewrite the evaluator — read `artifacts["answer"]`, return an `EvalOutput`. 4. Update the entry-point names in `pyproject.toml`. 5. `pip install --no-deps -e cookbooks/` and test with `rllm eval`.