By audn-ai
Red-team your AI voice and text agents with automated adversarial campaigns. Jailbreaks, prompt injection, data extraction, social engineering testing — powered by audn.ai.
Register and verify an AI agent on Audn.ai for security testing. Supports text API agents (OpenAI, Anthropic, custom LLMs) and voice agents (Twilio, Genesys, Amazon Connect). Use when the user wants to add a new agent to test.
Generate ISO 42001, SOC2, vulnerability, executive, or compliance reports from Audn.ai campaign results. Use when the user needs compliance evidence, audit documentation, or executive summaries of security testing.
Authenticate with Audn.ai for red-team security testing. Sets up OAuth login or API token. Run this before using any other audn skill. Use when the user needs to log in, authenticate, connect to Audn, or set up their API credentials.
Quick vulnerability scan of an AI agent endpoint. Runs only critical and high-severity attack scenarios for a fast security check. Use for a rapid smoke test before a full red-team audit.
Run a full adversarial red-team security audit against any AI agent (text or voice). Registers the agent, selects attack scenarios, executes a campaign, monitors progress, retrieves results, and generates a vulnerability report. Use when the user wants to security-test, red-team, pen-test, or audit an LLM or voice agent.
External network access
Connects to servers outside your machine
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Red-team your AI agents from any coding IDE. Automated adversarial security testing for voice and text AI agents — jailbreaks, prompt injection, data extraction, social engineering, compliance reporting.
Works with Claude Code, Cursor, Codex, Windsurf, Cline, Amp, and 40+ other coding agents.
Audn.ai account — sign up free
npx skills add audn-ai/skills --all -a claude-code
In Claude Code, run:
/audn-login
This registers the Audn.ai MCP server. Then:
/exit or Ctrl+C, then run claude again (MCP servers load at startup)/audn-login again/mcp → select audn-redteam → click "Authenticate" → log in via browserAfter login, tokens are stored in your system keychain and refresh automatically. No API keys needed.
Not using Claude Code? Generate a token at audn.ai/dashboard/settings, then:
export AUDN_API_TOKEN="your-bearer-token"
/audn-red https://api.openai.com/v1/chat/completions
That's it. Audn registers your agent, selects attack scenarios, runs the campaign, and delivers a full vulnerability report.
# Single skill
npx skills add audn-ai/skills --skill audn-red -a claude-code
# All skills
npx skills add audn-ai/skills --all -a claude-code
# Multiple agents at once
npx skills add audn-ai/skills --skill audn-red -a claude-code -a cursor -a codex
# Global install (available in all projects)
npx skills add audn-ai/skills --skill audn-red -a claude-code -g
# Claude Code plugin mode
claude --plugin-dir ./ # from this repo
audn-login — Authenticate with Audn.aiSet up authentication. On Claude Code, triggers OAuth browser login via the bundled MCP server. On other agents, guides API token setup.
/audn-login
audn-red — Full Red-Team AuditEnd-to-end adversarial security campaign. Registers your agent, verifies connectivity, selects attack scenarios from the library, creates and executes a campaign, monitors progress, retrieves results, checks for vulnerabilities, and generates a report.
/audn-red https://api.openai.com/v1/chat/completions
/audn-red https://api.anthropic.com/v1/messages
/audn-red https://your-internal-api.com/chat
Output: Security grade (A-F), vulnerability table, attack category breakdown, remediation recommendations.
audn-quick-scan — Fast Vulnerability ScanRuns only critical and high-severity scenarios for a rapid smoke test. Takes minutes instead of a full audit.
/audn-quick-scan https://api.openai.com/v1/chat/completions
audn-compliance-report — Compliance ReportsGenerate ISO 42001, SOC2, vulnerability, executive, or general compliance reports from completed campaign data.
/audn-compliance-report <campaign-id> iso_42001
/audn-compliance-report <campaign-id> soc2
/audn-compliance-report <campaign-id> executive
audn-agent-setup — Register & Verify AgentsRegister text API agents (OpenAI, Anthropic, custom LLMs) or voice agents (Twilio, Genesys, Amazon Connect) for testing.
/audn-agent-setup https://api.openai.com/v1/chat/completions
/audn-agent-setup +14155551234
audn-scenario-builder — Custom Attack ScenariosDesign custom adversarial attack scenarios — prompt injection, jailbreak, data extraction, social engineering, emotional manipulation, role confusion.
/audn-scenario-builder
audn-vuln-monitor — Vulnerability DashboardList and triage vulnerabilities across all your agents. Severity ratings, categories, remediation recommendations, trend analysis.
/audn-vuln-monitor
| Attack Category | What It Tests |
|---|---|
| Prompt Injection | System prompt override, instruction hijacking, context manipulation |
| Jailbreak | Safety guardrail bypass, persona override, DAN-style attacks |
| Data Extraction | PII leakage, credential exposure, system prompt extraction |
| Social Engineering | Authority impersonation, urgency manipulation, trust exploitation |
| Emotional Manipulation | Sympathy exploitation, guilt tripping, rapport abuse |
| Role Confusion | Identity confusion, role-play exploitation, context switching |
These skills orchestrate the Audn.ai API:
npx claudepluginhub audn-ai/skills --plugin audn-redteamSecurity check + optimize skills for chat system prompts and agent pipelines, plus agent-security skills (check/optimize/meta-learning)
Self-audit AI agent, tool, and MCP-server code for security and reliability misconfigurations with Trustabl, the static analyzer for the OpenAI Agents SDK, Claude Agent SDK, Google ADK, and MCP. Ships two skills (trustabl-scan and trustabl-enrich) and a subagent (trustabl) that together form a scan → enrich → review → apply pipeline.
AGT governance hooks and MCP tools for Claude Code sessions
Generate comprehensive security audit reports
Cybersecurity skills for AI agents — code audit, cloud, recon, IR, AI security, and more
GoPlus AgentGuard — AI agent security guard. Blocks dangerous commands, prevents data leaks, protects secrets. 20 detection rules, runtime action evaluation, trust registry.