RAMPART Examples

Runnable showcases of RAMPART: the pytest-native safety and security testing framework for agentic AI applications.

Each subdirectory is a self-contained demo. Pick one, follow its README.md, and you'll have a red -> fix -> green walkthrough running in minutes. Each demo declares its own dependencies, manifest, surface, adapter, and tests; nothing here at the repo root is required at runtime.

Demos

Demo	What it shows
helpdesk-bot	Indirect prompt injection (XPIA) via a poisoned support ticket. Single-git apply red -> green walkthrough.
openclaw	Plugin supply-chain XPIA against OpenClaw. Malicious plugin delivers payloads via tool output or tool description; auth-injecting reverse proxy keeps API keys off the sandbox.

Repository layout

rampart-examples/
├── README.md                # this file
├── pyproject.toml           # shared tooling config (ruff, ty, smoke-test pytest)
├── .pre-commit-config.yaml  # shared lint/format hooks (SHA-pinned)
├── .github/workflows/       # public CI: lint, smoke, patch round-trip
├── tests/                   # maintainer smoke tests (no LLM, no API keys)
└── <demo-name>/             # one folder per demo, fully self-contained
    ├── README.md            # walkthrough for THIS demo (the canonical artifact)
    ├── pyproject.toml       # demo's own dependencies and pytest config
    ├── <package_name>/      # the demo's installable Python package (flat layout)
    ├── tests/               # demo-local pytest fixtures + RAMPART tests
    └── mitigation.patch     # the unified-diff fix

For maintainers

The tests/ directory contains a fast smoke suite that exercises every demo's deterministic pieces (imports, manifests, surface lifecycle, predicates, mitigation-patch applicability) without making any LLM call. The repo is a uv workspace; uv sync installs every demo plus the maintainer toolchain.

uv sync
uv run pytest

Demo users running pytest from inside a demo folder never see this suite; each demo's own pyproject.toml scopes pytest to its own directory.

Public CI (lint + smoke + patch round-trip) runs in GitHub Actions on every PR. Live integration tests against real OpenAI / Azure OpenAI endpoints run in an internal Azure DevOps pipeline; see .azure-pipelines/README.md.

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit Contributor License Agreements.

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

Trademarks

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.