# Opfor

## Docs

- [Browser extension](https://docs.agentopfor.ai/browser-extension/overview.md): Red-team any chat UI from your browser — no code required.
- [Autonomous mode](https://docs.agentopfor.ai/cli/autonomous.md): Let an AI agent plan and run a full assessment from a single objective.
- [CLI overview](https://docs.agentopfor.ai/cli/overview.md): Install opfor and create a scan config.
- [Running scans](https://docs.agentopfor.ai/cli/running-scans.md): Fire attacks, judge responses, and read the report.
- [Authoring evaluators](https://docs.agentopfor.ai/evaluators/authoring.md): Add a new evaluator — no TypeScript needed.
- [Evaluators & suites](https://docs.agentopfor.ai/evaluators/overview.md): What opfor tests for, and how to choose.
- [Evaluator reference](https://docs.agentopfor.ai/evaluators/reference.md): Every evaluator and suite, with OWASP mappings.
- [Core concepts](https://docs.agentopfor.ai/get-started/concepts.md): The handful of ideas behind every opfor scan.
- [Quickstart](https://docs.agentopfor.ai/get-started/quickstart.md): Run your first red-team scan and get a report.
- [Introduction](https://docs.agentopfor.ai/index.md): Open-source adversary emulation for AI agents, LLM apps, and MCP servers.
- [MCP server](https://docs.agentopfor.ai/mcp-server/overview.md): Run opfor as an MCP server so AI assistants can drive it from chat.
- [CLI commands](https://docs.agentopfor.ai/reference/cli-commands.md): Every opfor command and flag.
- [Config reference](https://docs.agentopfor.ai/reference/config.md): Every field in an opfor config file.
- [LLM providers](https://docs.agentopfor.ai/reference/providers.md): Supported attacker/judge LLM providers and their env vars.
- [SDK](https://docs.agentopfor.ai/sdk/overview.md): Run opfor scans programmatically from your own TypeScript or Node code.
- [Skills](https://docs.agentopfor.ai/skills/overview.md): Drive opfor from your AI coding agent with slash commands.
- [Trace-aware testing](https://docs.agentopfor.ai/telemetry/overview.md): Let the judge see tool calls and retrievals, not just the reply.

## Optional

- [GitHub](https://github.com/KeyValueSoftwareSystems/agent-opfor)