redprobe 0.1.4__tar.gz → 0.1.6__tar.gz

{redprobe-0.1.4 → redprobe-0.1.6}/CONTRIBUTING.md

@@ -161,17 +161,55 @@ To run a subset of tests:
 pytest tests.test_redprobe
 ```
 
-## Deploying
+## Releasing to PyPI
 
-A reminder for the maintainers on how to deploy. Make sure all your changes are committed (including an entry in HISTORY.md). Then run:
+### First-Time Setup (One-Time)
 
-```sh
-bump2version patch # possible: major / minor / patch
-git push
-git push --tags
-```
+Before you can publish, set up PyPI trusted publishing:
+
+1. Create a PyPI account at [pypi.org](https://pypi.org) if you don't have one
+2. Go to your PyPI account > Publishing > Add a new pending publisher
+3. Fill in the form:
+   - **PyPI Project Name:** `redprobe`
+   - **Owner:** Your GitHub username or organization
+   - **Repository name:** `redprobe`
+   - **Workflow name:** `publish.yml`
+   - **Environment name:** Leave blank
+4. Push your first tag to trigger the publish
+
+### Every Release
+
+Releases are automated via GitHub Actions. When you push a tag, the workflow builds and publishes to PyPI using trusted publishing.
+
+1. Make sure all changes are committed and pushed
+2. Bump the version:
+
+   ```sh
+   uv version --bump patch  # or: minor, major
+   ```
+
+3. Commit the version bump:
+
+   ```sh
+   git add pyproject.toml uv.lock
+   git commit -m "Release X.Y.Z"
+   git push
+   ```
+
+4. Tag and push (triggers the publish):
+
+   ```sh
+   just tag
+   ```
+
+   Or manually:
+
+   ```sh
+   git tag -a vX.Y.Z -m "Release vX.Y.Z"
+   git push origin vX.Y.Z
+   ```
 
-You can set up a [GitHub Actions workflow](https://docs.github.com/en/actions/use-cases-and-examples/building-and-testing/building-and-testing-python#publishing-to-pypi) to automatically deploy your package to PyPI when you push a new tag.
+5. GitHub Actions publishes to PyPI automatically. Check the Actions tab to verify.
 
 ## Code of Conduct
 

{redprobe-0.1.4 → redprobe-0.1.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: redprobe
-Version: 0.1.4
+Version: 0.1.6
 Summary: A defensive security tool for hardening AI systems. Define YAML-based test cases to systematically probe LLMs for jailbreaks, prompt injections, biases, harmful content generation, data leakage, and policy violations before attackers find them. Compatible with any OpenAI-style API endpoint.
 Author-email: "Audrey M. Roy Greenfeld" <audrey@feldroy.com>
 Maintainer-email: "Audrey M. Roy Greenfeld" <audrey@feldroy.com>
@@ -24,7 +24,7 @@ Dynamic: license-file
 
 A defensive security tool for hardening AI systems. Define YAML-based test cases to systematically probe LLMs for jailbreaks, prompt injections, biases, harmful content generation, data leakage, and policy violations before attackers find them. Compatible with any OpenAI-style API endpoint.
 
-![RedProbe CLI](docs/cli.png)
+![RedProbe CLI](https://i.imgur.com/IAT0f7o.png)
 
 > **For authorized security testing only.** You must only test systems you own or have written permission to test. See [Responsible Use](#responsible-use) below.
 
@@ -351,6 +351,22 @@ uvx redprobe run probes/ \
 
 RedProbe works with any API that implements the OpenAI chat completions format (`/v1/chat/completions`). Set the base URL and model accordingly.
 
+## RedProbe vs PyRIT
+
+| Aspect | RedProbe | [PyRIT](https://github.com/Azure/PyRIT) |
+|--------|----------|-------|
+| **Complexity** | Simple CLI, run with `uvx redprobe` | Full framework requiring Python setup and code |
+| **Learning Curve** | Minutes: write YAML, run command | Hours/days: learn Python API, orchestrators, converters |
+| **Probe Definition** | YAML files with patterns | Python code with attack strategies |
+| **Target** | Any OpenAI-compatible API | Multi-modal, multi-platform (Azure, Hugging Face, etc.) |
+| **Detection** | Regex/string pattern matching | LLM-based scoring, custom scorers |
+| **Automation** | Run probes, get results | Multi-turn conversations, attack chaining, prompt mutation |
+| **Use Case** | Quick safety validation, CI/CD checks | Deep red teaming operations, research |
+
+**Use RedProbe** for quick safety checks, CI/CD integration, or testing specific prompts with minimal setup.
+
+**Use PyRIT** for extensive multi-day red teaming, multi-turn attack strategies, or deep security research.
+
 ## License
 
 BUSL 1.1. See [RESPONSIBLE_USE.md](RESPONSIBLE_USE.md) for usage guidelines.

{redprobe-0.1.4 → redprobe-0.1.6}/README.md

@@ -2,7 +2,7 @@
 
 A defensive security tool for hardening AI systems. Define YAML-based test cases to systematically probe LLMs for jailbreaks, prompt injections, biases, harmful content generation, data leakage, and policy violations before attackers find them. Compatible with any OpenAI-style API endpoint.
 
-![RedProbe CLI](docs/cli.png)
+![RedProbe CLI](https://i.imgur.com/IAT0f7o.png)
 
 > **For authorized security testing only.** You must only test systems you own or have written permission to test. See [Responsible Use](#responsible-use) below.
 
@@ -329,6 +329,22 @@ uvx redprobe run probes/ \
 
 RedProbe works with any API that implements the OpenAI chat completions format (`/v1/chat/completions`). Set the base URL and model accordingly.
 
+## RedProbe vs PyRIT
+
+| Aspect | RedProbe | [PyRIT](https://github.com/Azure/PyRIT) |
+|--------|----------|-------|
+| **Complexity** | Simple CLI, run with `uvx redprobe` | Full framework requiring Python setup and code |
+| **Learning Curve** | Minutes: write YAML, run command | Hours/days: learn Python API, orchestrators, converters |
+| **Probe Definition** | YAML files with patterns | Python code with attack strategies |
+| **Target** | Any OpenAI-compatible API | Multi-modal, multi-platform (Azure, Hugging Face, etc.) |
+| **Detection** | Regex/string pattern matching | LLM-based scoring, custom scorers |
+| **Automation** | Run probes, get results | Multi-turn conversations, attack chaining, prompt mutation |
+| **Use Case** | Quick safety validation, CI/CD checks | Deep red teaming operations, research |
+
+**Use RedProbe** for quick safety checks, CI/CD integration, or testing specific prompts with minimal setup.
+
+**Use PyRIT** for extensive multi-day red teaming, multi-turn attack strategies, or deep security research.
+
 ## License
 
 BUSL 1.1. See [RESPONSIBLE_USE.md](RESPONSIBLE_USE.md) for usage guidelines.

{redprobe-0.1.4 → redprobe-0.1.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "redprobe"
-version = "0.1.4"
+version = "0.1.6"
 description = "A defensive security tool for hardening AI systems. Define YAML-based test cases to systematically probe LLMs for jailbreaks, prompt injections, biases, harmful content generation, data leakage, and policy violations before attackers find them. Compatible with any OpenAI-style API endpoint."
 readme = "README.md"
 authors = [

{redprobe-0.1.4 → redprobe-0.1.6}/src/redprobe.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: redprobe
-Version: 0.1.4
+Version: 0.1.6
 Summary: A defensive security tool for hardening AI systems. Define YAML-based test cases to systematically probe LLMs for jailbreaks, prompt injections, biases, harmful content generation, data leakage, and policy violations before attackers find them. Compatible with any OpenAI-style API endpoint.
 Author-email: "Audrey M. Roy Greenfeld" <audrey@feldroy.com>
 Maintainer-email: "Audrey M. Roy Greenfeld" <audrey@feldroy.com>
@@ -24,7 +24,7 @@ Dynamic: license-file
 
 A defensive security tool for hardening AI systems. Define YAML-based test cases to systematically probe LLMs for jailbreaks, prompt injections, biases, harmful content generation, data leakage, and policy violations before attackers find them. Compatible with any OpenAI-style API endpoint.
 
-![RedProbe CLI](docs/cli.png)
+![RedProbe CLI](https://i.imgur.com/IAT0f7o.png)
 
 > **For authorized security testing only.** You must only test systems you own or have written permission to test. See [Responsible Use](#responsible-use) below.
 
@@ -351,6 +351,22 @@ uvx redprobe run probes/ \
 
 RedProbe works with any API that implements the OpenAI chat completions format (`/v1/chat/completions`). Set the base URL and model accordingly.
 
+## RedProbe vs PyRIT
+
+| Aspect | RedProbe | [PyRIT](https://github.com/Azure/PyRIT) |
+|--------|----------|-------|
+| **Complexity** | Simple CLI, run with `uvx redprobe` | Full framework requiring Python setup and code |
+| **Learning Curve** | Minutes: write YAML, run command | Hours/days: learn Python API, orchestrators, converters |
+| **Probe Definition** | YAML files with patterns | Python code with attack strategies |
+| **Target** | Any OpenAI-compatible API | Multi-modal, multi-platform (Azure, Hugging Face, etc.) |
+| **Detection** | Regex/string pattern matching | LLM-based scoring, custom scorers |
+| **Automation** | Run probes, get results | Multi-turn conversations, attack chaining, prompt mutation |
+| **Use Case** | Quick safety validation, CI/CD checks | Deep red teaming operations, research |
+
+**Use RedProbe** for quick safety checks, CI/CD integration, or testing specific prompts with minimal setup.
+
+**Use PyRIT** for extensive multi-day red teaming, multi-turn attack strategies, or deep security research.
+
 ## License
 
 BUSL 1.1. See [RESPONSIBLE_USE.md](RESPONSIBLE_USE.md) for usage guidelines.