# llm-redteam-specialist

Subagent designs adversarial probe suites, runs jailbreak evaluations, and drafts compliance evidence packages for deployed language models.

- subagent · `type=agent` · Claude Code · 139 installs
- Author: davila7
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/davila7/claude-code-templates/tree/HEAD/cli-tool/components/agents/security/llm-redteam-specialist.md
- Updated: 2026-10-24
- HTML: https://affpapa.org/en/nevibecode/agents/llm-redteam-specialist

## What it does

The subagent structures adversarial assessments for both cloud APIs and local deployments using Ollama, vLLM, or llama.cpp. It covers direct jailbreak families, prompt leaking, context manipulation, and indirect injections embedded in retrieval contexts.

It operates with Read, Grep, Glob, and Bash to inspect repositories and configure test runners. Grading relies on deterministic keyword or regex patterns in restricted enclaves, or calibrated model-as-judge setups when external networks are permitted. Findings map directly to NIST AI RMF MEASURE-2.7, EU AI Act Article 15, and the OWASP LLM Top 10 taxonomy.

Without network access, the agent defaults entirely to offline rubrics and avoids remote scoring APIs. It can integrate references to tooling such as Garak, Promptfoo, and PyRIT.

## Install

### Claude Code (native platform)

1. **claude-code-templates**
```bash
npx claude-code-templates@latest --agent security/llm-redteam-specialist
```

2. **Manually: put the files in place**
- File: `.claude/agents/llm-redteam-specialist.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/llm-redteam-specialist.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/llm-redteam-specialist.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/davila7/claude-code-templates/tree/HEAD/cli-tool/components/agents/security/llm-redteam-specialist.md

## When to use it
- Build an evidence bundle mapped to EU AI Act Article 15 and NIST AI RMF
- Design an offline prompt injection test harness for Ollama endpoints
- Audit retrieval pipelines against indirect injection through external documents
- Generate an adversarial test report aligned with the OWASP LLM Top 10

## Example prompts
- Build an air-gapped test suite to check our vLLM deployment for indirect prompt injection
- Generate a prompt injection compliance report mapped to OWASP LLM Top 10 for enterprise security review
- Audit our retrieval pipeline prompts against system instruction extraction and obfuscation attacks

## Similar
- [code-reviewer](https://affpapa.org/en/nevibecode/agents/code-reviewer.md) — subagent · 31050 installs
- [security-auditor](https://affpapa.org/en/nevibecode/agents/security-auditor.md) — subagent · 10718 installs
- [test-engineer](https://affpapa.org/en/nevibecode/agents/test-engineer.md) — subagent · 9715 installs
- [code-reviewer](https://affpapa.org/en/nevibecode/agents/voltagent-community-code-reviewer.md) — subagent · 31050 installs
- [security-auditor](https://affpapa.org/en/nevibecode/agents/voltagent-community-security-auditor.md) — subagent · 10718 installs
- [devops-engineer](https://affpapa.org/en/nevibecode/agents/devops-engineer.md) — subagent · 7392 installs

From the same set: [davila7/claude-code-templates](https://affpapa.org/en/nevibecode/by/davila7/claude-code-templates.md) — 804 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md