ADVERSA - Harness de Red-Team para LLMs
cognis-digital/adversa
Red-team harness MCP-nativo para LLMs — OWASP + MITRE em segundos
Ver no GitHubMCP server para teste de segurança ofensivo em modelos de linguagem, cobrindo OWASP LLM Top 10 e ataques MITRE ATLAS. Auto-hospedável, pronto para CI/CD e acionável por agentes, sem infraestrutura pesada. Entrega resultados priorizados em tabelas, JSON ou SARIF.
Instalação
npx convez-ai cognis-digital/adversa1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Teste de segurança ofensivo em modelos e aplicações LLM
- Gate automático de segurança em pipelines CI/CD
- Auditoria de vulnerabilidades em prompts e chains de agentes
- Validação de mitigações contra ataques conhecidos em LLM
Veredito

Ferramenta especializada e madura para adversarial testing em LLMs. Essencial para times que precisam de testes de segurança ofensivos estruturados e integrados em fluxos de CI — mas requer compreensão técnica sólida de segurança LLM e infraestrutura.
README
A tradução para o português deste README ainda está na fila.
ADVERSA
LLM red-team harness — OWASP LLM Top 10 + MITRE ATLAS attack packs
AI Security & Governance — securing LLMs, agents, and the MCP supply chain.
pip install cognis-adversa
adversa scan . # → prioritized findings in seconds
Contents
- Why adversa? · Features · Quick start · Example · Architecture · AI stack · How it compares · Integrations · Install anywhere · Related · Contributing
Why adversa?
LLM red-team harness — OWASP LLM Top 10 + MITRE ATLAS attack packs — without standing up heavyweight infrastructure.
adversa is single-purpose, scriptable, and self-hostable: point it at a target, get prioritized results in the format your workflow already speaks (table · JSON · SARIF), gate CI on it, and let agents drive it over MCP.
Features
- ✅ Severity Rank
- ✅ Builtin Probes
- ✅ Load Probes
- ✅ Detect Success
- ✅ Run Probes
- ✅ Transcript Target
- ✅ Load Transcript
- ✅ Runs on Linux/macOS/Windows · Docker · devcontainer
- ✅ Ports in Python, JavaScript, Go, and Rust (
ports/)
Quick start
pip install cognis-adversa
adversa --version
adversa scan . # scan current project
adversa scan . --format json # machine-readable
adversa scan . --fail-on high # CI gate (non-zero exit)
Example
$ adversa scan .
[HIGH ] ADV-001 example finding (./src/app.py)
[MEDIUM ] ADV-002 another signal (./config.yaml)
2 findings · risk score 5 · 38ms
Architecture
flowchart LR
A[Input: file / dir / API] --> B[Collectors]
B --> C[Rules / Analyzers]
C --> D[Scorer]
D --> E{Reporters}
E --> F[Table]
E --> G[JSON / SARIF]
E --> H[MCP tool -. drives .-> AI agents]
Use it from any AI stack
adversa is interoperable with every popular way of using AI:
- MCP server —
adversa mcp(Claude Desktop, Cursor, Cognis.Studio, uncensored-fleet) - OpenAI-compatible / JSON — pipe
adversa scan . --format jsoninto any agent or LLM - LangChain · CrewAI · AutoGen · LlamaIndex — wrap the CLI/JSON as a tool in one line
- CI / scripts — exit codes + SARIF for non-AI pipelines
How it compares
| Cognis adversa | leondz | |
|---|---|---|
| Self-hostable, no account | ✅ | varies |
| Single command, zero config | ✅ | ⚠️ |
| JSON + SARIF for CI | ✅ | varies |
| MCP-native (AI agents) | ✅ | ❌ |
| Polyglot ports (JS/Go/Rust) | ✅ | ❌ |
| Open license | ✅ COCL | varies |
Built in the spirit of leondz/garak, re-framed the Cognis way. Missing a credit? Open a PR.
Integrations
Pipes into your stack: SARIF for code-scanning, JSON for anything, an MCP server (adversa mcp) for AI agents, and a webhook forwarder for SIEM/Slack/Jira. See docs/INTEGRATIONS.md.
Install — every way, every platform
pip install "git+https://github.com/cognis-digital/adversa.git" # pip (works today)
pipx install "git+https://github.com/cognis-digital/adversa.git" # isolated CLI
uv tool install "git+https://github.com/cognis-digital/adversa.git" # uv
pip install cognis-adversa # PyPI (when published)
docker run --rm ghcr.io/cognis-digital/adversa:latest --help # Docker
brew install cognis-digital/tap/adversa # Homebrew tap
curl -fsSL https://raw.githubusercontent.com/cognis-digital/adversa/main/install.sh | sh
| Linux | macOS | Windows | Docker | Cloud |
|---|---|---|---|---|
scripts/setup-linux.sh | scripts/setup-macos.sh | scripts/setup-windows.ps1 | docker run ghcr.io/cognis-digital/adversa | DEPLOY.md (AWS/Azure/GCP/k8s) |
Related Cognis tools
aegis— AI Agent Permission & Access Auditor — surfaces the lethal trifecta of credentials + injection + reachpromptmirror— Prompt-injection & indirect-injection scanner for any LLM context inputledgermind— Local LLM cost & token forensics proxy with anomaly detectionguardpost— Runtime agent firewall — PII redaction, rate limits, policy enforcementhallumark— LLM hallucination & grounding auditor for RAG systemsaicard— Auto-generated NIST AI RMF / EU AI Act Annex IV model & system cards
Explore the suite → 🗂️ all 170+ tools · ⭐ awesome-cognis · 🔗 cognis-sources · 🤖 uncensored-fleet · 🧠 hermes
Contributing
PRs, new rules, and demo scenarios are welcome under the collaboration-pull model — see CONTRIBUTING.md and SECURITY.md.
⭐ If
adversasaved you time, star it — it genuinely helps others find it.
License
Source-available under the Cognis Open Collaboration License (COCL) v1.0 — free for personal, internal-evaluation, research, and educational use; commercial / production use requires a license (licensing@cognis.digital). See LICENSE.