Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

SuperRed

rdi.berkeley.edu

  • Free plan
  • Linux
6.9 Club score Visit rdi.berkeley.edu

What the maker publishes

Attacks and benchmarks
The module catalogue lists 35 modules: 21 attackers, 6 targets, and 8 benchmarks. source
Composable components
It treats the attacker, system under test, and benchmark as separate interchangeable modules coordinated by a controller. source
Composable modules
It keeps attackers, systems under test, and security claims as interchangeable modules coordinated by a controller. source
Evaluation reports
Runs record model, cost, and success rate; the framework provides live progress and browsable reports with per-task details and trajectories. source
Example modules
Listed modules include PAIR, TAP, AutoDAN-Turbo, Crescendo, AgentVigil, HarmBench, AgentDojo, and DecodingTrust-Agent. source
Installation
The guide installs the framework with pip install superred and describes it as a Python framework whose guide assumes familiarity with Python and asyncio. source
Integration
The getting-started example connects to any LiteLLM-compatible model endpoint using a base URL and API key. source
Intended audience
The getting-started guide is for people who can read Python and have seen asyncio, and PyPI lists Science/Research as an intended audience. source
Intended users
The project describes use by red-teamers, system builders, and evaluators, and its guide covers wrapping systems, writing attackers, defining success criteria, and running evaluations. source
License and maturity
PyPI lists the package under the MIT license and classifies its development status as Alpha. source
Maker
The site says SuperRed was made at the University of California, Berkeley. source
Metrics and reports
Runs record model, cost, and success rate, with a live terminal dashboard and a web report for results. source
Model endpoints
The example target uses a LiteLLM-compatible endpoint with a base URL and API key; the guide says most LLM-driven attackers also call models through LiteLLM. source
Parallel runs
The framework can run multiple threat models in parallel, each against its own system instance. source
Purpose
SuperRed is an open-source framework for red-teaming chatbots, agents, and assistants by testing whether attacks can violate security properties. source
Requirements
The package requires Python 3.11 through 3.13; Python 3.14 is not supported. source
Scale
The controller can run many threat models in parallel, each against its own system instance. source
Security controls
The controller filters attacker-visible and injectable surfaces according to security-domain scopes and can cap attacker model spend per task. source
Security scope
The controller filters which controllables, observables, trajectory entries, and evaluation sub-scores the optimizer can access according to the configured scope. source
Sensitive results
Persisted trajectories are not scrubbed and may contain jailbreaks, planted secrets, and exfiltrated content; API keys and API base URLs are not written in the serialized LLM configuration. source
Sensitive results
Persisted trajectories are not scrubbed and may contain jailbreaks, planted secrets, and exfiltrated content, so the maker says to treat the results folder as sensitive. source
Supported targets
Targets can wrap fixed-response fixtures, simulated environments, sandboxes, or live deployments, and the guide recommends staging or throwaway instances for real systems. source
Target types
Targets can wrap fixed-response fixtures, simulated environments, sandboxes, or live deployments. source
Threat models
Each run can define attacker model, per-task budget, trust-boundary access, and whether benchmark feedback is visible to the attacker. source

Checked against the maker's pages on 4 Oct 2026.