diff --git a/SETUP.MD b/SETUP.MD
new file mode 100644
index 00000000..d7b750a8
--- /dev/null
+++ b/SETUP.MD
@@ -0,0 +1,141 @@
+# SETUP.MD
+
+This file documents how to provision a clean development environment for the five packages in this repo (`uipath-agent-framework`, `uipath-google-adk`, `uipath-llamaindex`, `uipath-openai-agents`, `uipath-pydantic-ai`), run the build, execute the tests, and validate a sample code change end-to-end. It is intended both as a quick reference for human contributors and as a structured guide for automated environment-setup tooling.
+
+## Prerequisites
+
+- Python 3.11+
+- [uv](https://docs.astral.sh/uv/) 0.5+
+
+### Supported platforms
+
+`uv` is shell- and OS-agnostic, so the commands below run unchanged on every supported platform:
+
+- [x] Linux
+- [x] Windows
+- [x] macOS
+
+## Environment Variables
+
+None required for environment setup, build, or unit tests. The suites under the `Test` section run fully offline and require no external authentication.
+
+> **All commands below must be run from the repository root.** The `uv --directory packages/<name>` invocations resolve each subpackage relative to the current working directory. The first line of `## Setup` enforces this by `cd`-ing to the git root.
+
+## Setup
+
+```bash
+cd "$(git rev-parse --show-toplevel)"
+python3 -m pip install --upgrade uv
+
+# Sync all five packages (each is independent)
+uv --directory packages/uipath-agent-framework sync --all-extras
+uv --directory packages/uipath-google-adk sync --all-extras
+uv --directory packages/uipath-llamaindex sync --all-extras
+uv --directory packages/uipath-openai-agents sync --all-extras
+uv --directory packages/uipath-pydantic-ai sync --all-extras
+```
+
+## Verify Setup
+
+```bash
+uv --version
+uv --directory packages/uipath-pydantic-ai run python --version
+uv --directory packages/uipath-agent-framework run python -c "import uipath_agent_framework; print('uipath-agent-framework ok')"
+uv --directory packages/uipath-google-adk run python -c "import uipath_google_adk; print('uipath-google-adk ok')"
+uv --directory packages/uipath-llamaindex run python -c "import uipath_llamaindex; print('uipath-llamaindex ok')"
+uv --directory packages/uipath-openai-agents run python -c "import uipath_openai_agents; print('uipath-openai-agents ok')"
+uv --directory packages/uipath-pydantic-ai run python -c "import uipath_pydantic_ai; print('uipath-pydantic-ai ok')"
+```
+
+## Build
+
+N/A
+
+## Test
+
+```bash
+uv --directory packages/uipath-agent-framework run pytest
+uv --directory packages/uipath-google-adk run pytest
+uv --directory packages/uipath-llamaindex run pytest
+uv --directory packages/uipath-openai-agents run pytest
+uv --directory packages/uipath-pydantic-ai run pytest
+```
+
+## Sample Code Change
+
+### The change
+
+Add a new `agent_count` property to `PydanticAiConfig` in `packages/uipath-pydantic-ai/src/uipath_pydantic_ai/runtime/config.py`, immediately after the existing `entrypoint` property:
+
+```python
+@property
+def agent_count(self) -> int:
+    """Number of agents defined in the configuration."""
+    return len(self.agents)
+```
+
+Then create `packages/uipath-pydantic-ai/tests/test_config_agent_count.py` with two pytest tests:
+
+```python
+"""Tests for PydanticAiConfig.agent_count."""
+
+import json
+from pathlib import Path
+
+from uipath_pydantic_ai.runtime.config import PydanticAiConfig
+
+
+def test_agent_count_single(tmp_path: Path) -> None:
+    config_path = tmp_path / "pydantic_ai.json"
+    config_path.write_text(json.dumps({"agents": {"main": "main:agent"}}))
+    cfg = PydanticAiConfig(str(config_path))
+    assert cfg.agent_count == 1
+
+
+def test_agent_count_multiple(tmp_path: Path) -> None:
+    config_path = tmp_path / "pydantic_ai.json"
+    config_path.write_text(
+        json.dumps(
+            {
+                "agents": {
+                    "alpha": "alpha:agent",
+                    "beta": "beta:agent",
+                    "gamma": "gamma:agent",
+                }
+            }
+        )
+    )
+    cfg = PydanticAiConfig(str(config_path))
+    assert cfg.agent_count == 3
+```
+
+### Verification
+
+```bash
+uv --directory packages/uipath-pydantic-ai run pytest tests/test_config_agent_count.py -v
+```
+
+## Test with a real UiPath Coded Agent
+
+The unit tests above are necessary but not sufficient — they don't exercise the package end-to-end through a real agent. The flow below validates changes against a live runtime:
+
+1. Apply the code changes locally.
+2. Run the unit tests (see the `Sample Code Change` section above).
+3. Scaffold a coded UiPath agent (PydanticAI / OpenAI / Google ADK / LlamaIndex / Agent Framework, matching the package you changed) that exercises the changed code path.
+4. In the downstream project's `pyproject.toml`, add this local library as an editable dependency (substitute the package you changed):
+
+   ```toml
+   [tool.uv.sources]
+   uipath-pydantic-ai = { path = "../path/to/uipath-integrations-python/packages/uipath-pydantic-ai", editable = true }
+   ```
+
+5. Exercise the new behavior end-to-end:
+
+   ```bash
+   uv run uipath run <agent-name> --input '{...}'
+   ```
+
+6. (Optional) Open a PR and apply the `build:dev` label — this publishes the development version to Test PyPI.
+7. The PR description is updated automatically with instructions for pointing the downstream agent at the Test PyPI dev version.
+8. Push the dev version to UiPath with [`uipath push`](https://uipath.github.io/uipath-python/cli/#push), then deploy it to Orchestrator or Studio Web with [`uipath deploy`](https://uipath.github.io/uipath-python/cli/#deploy), and run it in cloud to confirm the changes behave correctly against the real platform.
+9. Once validation is done, close the dev PR — these PRs are not meant to be merged; their only purpose was to publish a Test PyPI build for end-to-end validation.