RAMPART
RAMPART is a free, open-source framework for writing and running safety and security tests for agentic AI applications.
What RAMPART is
RAMPART is a free, open-source framework for writing and running safety and security tests for agentic AI applications. It works within pytest: users connect an agent through an adapter, while RAMPART provides execution strategies, evaluators, test integration, and reporting. Its attack tests mark detected attacks as UNSAFE; behavior probes mark detected expected behavior as SAFE. The documented built-in attack checks cross-prompt injection, including whether malicious content in a data source can affect an agent when retrieved. The built-in probe checks expected responses, tool use, or side effects, making probes useful for regression testing. Results include a SAFE, UNSAFE, UNDETERMINED, or ERROR status, plus trace and evaluation details. RAMPART registers automatically as a pytest plugin and provides harm and trial markers. It creates structured JSON reports, supports custom report sinks, and can combine results from parallel pytest-xdist workers. Installation requires Python 3.11 or later and uses uv or pip on Linux, macOS, or Windows. PyRIT is a dependency; an optional OneDrive extra supports placing XPIA payloads in OneDrive.
Good
- Free and MIT-licensed.
- Includes cross-prompt injection attack testing.
- Behavior probes can check responses, tool use, and side effects.
- Supports structured JSON reports and custom report sinks.
- Combines reports from parallel pytest-xdist workers.
Watch out
- Requires Python 3.11 or later.
- PyRIT is installed as a dependency.
- OneDrive payload support requires an optional extra.
- Teardown-only results are excluded from xdist report streaming.
Who it suits
RAMPART suits developers who want to write and run safety tests for AI agents within pytest workflows. It is relevant to teams checking prompt-injection risks or expected agent behavior.
Verdict
RAMPART brings agent safety checks into pytest, with attack tests, behavior probes, and structured results. Teams should account for its Python requirement and the documented limits on xdist reporting.
The facts
Each one read from RAMPART's own pages, with the date it was checked.
What it does
- Attack and probe tests
- Its two test categories are attacks, where a detected attack means UNSAFE, and probes, where detected expected behavior means SAFE.checked 8 Oct 2026
- Attack testing
- Its XPIA tests check whether malicious content planted in data sources can manipulate an agent when retrieved.checked 9 Oct 2026
- Behavior probes
- Behavioral probes check for expected responses, tool use, or side effects and are described as useful for regression testing.checked 9 Oct 2026
- Evaluation output
- Results include a safety status of SAFE, UNSAFE, UNDETERMINED, or ERROR, along with trace and evaluation details.checked 9 Oct 2026
- Included tests
- The documented built-in attack is cross-prompt injection (XPIA), and the documented probe verifies expected agent behavior.checked 8 Oct 2026
- Install requirements
- The installation guide requires Python 3.11 or later and supports installation with uv or pip.checked 9 Oct 2026
- Integration model
- Users provide an adapter connecting their agent to RAMPART, which supplies execution strategies, evaluators, pytest integration, and reporting.checked 8 Oct 2026
- Intended users
- The project describes its framework as a developer-friendly way to write and run safety and security tests for AI agents.checked 9 Oct 2026
- License
- The GitHub repository identifies RAMPART as MIT-licensed.checked 9 Oct 2026
- Limitations
- The xdist documentation says results recorded only during fixture teardown are excluded from report streaming and mixed RAMPART versions across controller and workers are unsupported.checked 8 Oct 2026
- OneDrive
- An optional onedrive extra installs dependencies for using the built-in OneDriveSurface to place XPIA payloads in OneDrive.checked 8 Oct 2026
- Optional integration
- The optional onedrive extra adds dependencies for using the built-in OneDriveSurface to place XPIA payloads in OneDrive.checked 9 Oct 2026
- Parallel execution
- RAMPART supports pytest-xdist parallel runs and produces a unified report across worker processes.checked 8 Oct 2026
- Purpose
- RAMPART is a pytest-native safety testing framework for agentic AI applications.checked 9 Oct 2026
- PyRIT
- RAMPART uses PyRIT for LLM interaction, prompt normalization, and conversation memory, and installs PyRIT as a dependency.checked 8 Oct 2026
- PyRIT integration
- RAMPART uses PyRIT for LLM interaction, prompt normalization, and conversation memory, and installs PyRIT as a dependency.checked 9 Oct 2026
- pytest integration
- RAMPART registers as a pytest plugin automatically and provides harm and trial markers.checked 9 Oct 2026
- pytest plugin
- RAMPART registers automatically as a pytest plugin through the pytest11 entry point, without requiring conftest.py activation configuration.checked 8 Oct 2026
- Python requirement
- Installation requires Python 3.11 or later and uses uv or pip for package management.checked 8 Oct 2026
- Reporting
- RAMPART provides structured JSON reporting, including a built-in JSON file report sink and a hook for custom sinks.checked 8 Oct 2026
- Reporting and CI
- Built-in JSON report sinks and pytest-xdist support let teams collect reports for CI dashboards across parallel workers.checked 9 Oct 2026
- Security boundary
- For xdist worker serialization, RAMPART documents JSON-safe payloads, rejection of unknown schema versions, a default 16 MiB per-result cap, and ANSI escape stripping.checked 8 Oct 2026
- Supported installation systems
- The installation instructions provide commands for Linux, macOS, and Windows.checked 8 Oct 2026
The company
- Maker
- The documentation identifies Microsoft Corporation as its copyright holder.checked 8 Oct 2026
Alternatives to RAMPART
- 1
OpenSecureAI Scanner
Free planOpen sourceWebAPILinuxmacOS8.1From $49 a month - 2

PromptGuard
Free planWebAPIExtensionLinux7.9From $19 a month - 3

Promptfoo
Free planOpen sourceWebAPILinuxmacOS7.8Free plan - 4

Giskard
Free planOpen sourceWebAPILinuxSelf-hosted7.6Free plan - 5

AIRTA Red Team
Free planOpen sourceWebAPILinuxmacOS7.5Free plan - 6
CyberSecEval
Free planOpen sourceWebAPILinuxmacOS7.5Free plan
