Skip to content

Targets

LLM API targets for testing.

LLMTarget

Bases: ABC

Abstract base class for LLM targets.

Source code in src/rotalabs_redqueen/llm/targets.py 
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
class LLMTarget(ABC):
    """Abstract base class for LLM targets."""

    @abstractmethod
    async def query(self, prompt: str) -> TargetResponse:
        """Send a prompt to the LLM and get response.

        Args:
            prompt: The prompt to send

        Returns:
            Target response with content and metadata

        Raises:
            TargetError: If the query fails
            RateLimitError: If rate limited
        """
        pass

    @property
    @abstractmethod
    def name(self) -> str:
        """Human-readable name of this target."""
        pass

name: str abstractmethod property

Human-readable name of this target.

query(prompt: str) -> TargetResponse abstractmethod async

Send a prompt to the LLM and get response.

Parameters:

Name Type Description Default
prompt str

The prompt to send

 required

Returns:

Type Description
TargetResponse

Target response with content and metadata

Raises:

Type Description
TargetError

If the query fails
RateLimitError

If rate limited
Source code in src/rotalabs_redqueen/llm/targets.py 
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
@abstractmethod
async def query(self, prompt: str) -> TargetResponse:
    """Send a prompt to the LLM and get response.

    Args:
        prompt: The prompt to send

    Returns:
        Target response with content and metadata

    Raises:
        TargetError: If the query fails
        RateLimitError: If rate limited
    """
    pass

OpenAITarget

Bases: LLMTarget

OpenAI API target (GPT-4, etc.).

Source code in src/rotalabs_redqueen/llm/targets.py 
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
class OpenAITarget(LLMTarget):
    """OpenAI API target (GPT-4, etc.)."""

    def __init__(
        self,
        model: str = "gpt-4",
        api_key: str | None = None,
        max_tokens: int = 1000,
    ):
        """Initialize OpenAI target.

        Args:
            model: Model name (gpt-4, gpt-3.5-turbo, etc.)
            api_key: API key (or uses OPENAI_API_KEY env var)
            max_tokens: Maximum tokens in response
        """
        self.model = model
        self.api_key = api_key or os.environ.get("OPENAI_API_KEY")
        self.max_tokens = max_tokens

        if not self.api_key:
            raise ValueError("OpenAI API key required")

    @property
    def name(self) -> str:
        return f"openai:{self.model}"

    async def query(self, prompt: str) -> TargetResponse:
        """Query OpenAI API."""
        import time

        import httpx

        start = time.time()

        async with httpx.AsyncClient() as client:
            try:
                response = await client.post(
                    "https://api.openai.com/v1/chat/completions",
                    headers={
                        "Authorization": f"Bearer {self.api_key}",
                        "Content-Type": "application/json",
                    },
                    json={
                        "model": self.model,
                        "messages": [{"role": "user", "content": prompt}],
                        "max_tokens": self.max_tokens,
                    },
                    timeout=60.0,
                )

                if response.status_code == 429:
                    raise RateLimitError("OpenAI rate limit exceeded")
                response.raise_for_status()

                data = response.json()
                latency = (time.time() - start) * 1000

                return TargetResponse(
                    content=data["choices"][0]["message"]["content"],
                    model=self.model,
                    tokens_used=data.get("usage", {}).get("total_tokens", 0),
                    latency_ms=latency,
                )
            except httpx.HTTPStatusError as e:
                raise TargetError(f"OpenAI API error: {e}") from e
            except httpx.RequestError as e:
                raise NetworkError(f"Network error: {e}") from e

__init__(model: str = 'gpt-4', api_key: str | None = None, max_tokens: int = 1000)

Initialize OpenAI target.

Parameters:

Name Type Description Default
model str

Model name (gpt-4, gpt-3.5-turbo, etc.)

 'gpt-4'
api_key str | None

API key (or uses OPENAI_API_KEY env var)

 None
max_tokens int

Maximum tokens in response

 1000
Source code in src/rotalabs_redqueen/llm/targets.py 
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
def __init__(
    self,
    model: str = "gpt-4",
    api_key: str | None = None,
    max_tokens: int = 1000,
):
    """Initialize OpenAI target.

    Args:
        model: Model name (gpt-4, gpt-3.5-turbo, etc.)
        api_key: API key (or uses OPENAI_API_KEY env var)
        max_tokens: Maximum tokens in response
    """
    self.model = model
    self.api_key = api_key or os.environ.get("OPENAI_API_KEY")
    self.max_tokens = max_tokens

    if not self.api_key:
        raise ValueError("OpenAI API key required")

query(prompt: str) -> TargetResponse async

Query OpenAI API.

Source code in src/rotalabs_redqueen/llm/targets.py 
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
async def query(self, prompt: str) -> TargetResponse:
    """Query OpenAI API."""
    import time

    import httpx

    start = time.time()

    async with httpx.AsyncClient() as client:
        try:
            response = await client.post(
                "https://api.openai.com/v1/chat/completions",
                headers={
                    "Authorization": f"Bearer {self.api_key}",
                    "Content-Type": "application/json",
                },
                json={
                    "model": self.model,
                    "messages": [{"role": "user", "content": prompt}],
                    "max_tokens": self.max_tokens,
                },
                timeout=60.0,
            )

            if response.status_code == 429:
                raise RateLimitError("OpenAI rate limit exceeded")
            response.raise_for_status()

            data = response.json()
            latency = (time.time() - start) * 1000

            return TargetResponse(
                content=data["choices"][0]["message"]["content"],
                model=self.model,
                tokens_used=data.get("usage", {}).get("total_tokens", 0),
                latency_ms=latency,
            )
        except httpx.HTTPStatusError as e:
            raise TargetError(f"OpenAI API error: {e}") from e
        except httpx.RequestError as e:
            raise NetworkError(f"Network error: {e}") from e

AnthropicTarget

Bases: LLMTarget

Anthropic API target (Claude).

Source code in src/rotalabs_redqueen/llm/targets.py 
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
class AnthropicTarget(LLMTarget):
    """Anthropic API target (Claude)."""

    def __init__(
        self,
        model: str = "claude-sonnet-4-20250514",
        api_key: str | None = None,
        max_tokens: int = 1000,
    ):
        """Initialize Anthropic target.

        Args:
            model: Model name (claude-sonnet-4-20250514, etc.)
            api_key: API key (or uses ANTHROPIC_API_KEY env var)
            max_tokens: Maximum tokens in response
        """
        self.model = model
        self.api_key = api_key or os.environ.get("ANTHROPIC_API_KEY")
        self.max_tokens = max_tokens

        if not self.api_key:
            raise ValueError("Anthropic API key required")

    @property
    def name(self) -> str:
        return f"anthropic:{self.model}"

    async def query(self, prompt: str) -> TargetResponse:
        """Query Anthropic API."""
        import time

        import httpx

        start = time.time()

        async with httpx.AsyncClient() as client:
            try:
                response = await client.post(
                    "https://api.anthropic.com/v1/messages",
                    headers={
                        "x-api-key": self.api_key,
                        "anthropic-version": "2023-06-01",
                        "Content-Type": "application/json",
                    },
                    json={
                        "model": self.model,
                        "messages": [{"role": "user", "content": prompt}],
                        "max_tokens": self.max_tokens,
                    },
                    timeout=60.0,
                )

                if response.status_code == 429:
                    raise RateLimitError("Anthropic rate limit exceeded")
                response.raise_for_status()

                data = response.json()
                latency = (time.time() - start) * 1000

                return TargetResponse(
                    content=data["content"][0]["text"],
                    model=self.model,
                    tokens_used=data.get("usage", {}).get("input_tokens", 0)
                    + data.get("usage", {}).get("output_tokens", 0),
                    latency_ms=latency,
                )
            except httpx.HTTPStatusError as e:
                raise TargetError(f"Anthropic API error: {e}") from e
            except httpx.RequestError as e:
                raise NetworkError(f"Network error: {e}") from e

__init__(model: str = 'claude-sonnet-4-20250514', api_key: str | None = None, max_tokens: int = 1000)

Initialize Anthropic target.

Parameters:

Name Type Description Default
model str

Model name (claude-sonnet-4-20250514, etc.)

 'claude-sonnet-4-20250514'
api_key str | None

API key (or uses ANTHROPIC_API_KEY env var)

 None
max_tokens int

Maximum tokens in response

 1000
Source code in src/rotalabs_redqueen/llm/targets.py 
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
def __init__(
    self,
    model: str = "claude-sonnet-4-20250514",
    api_key: str | None = None,
    max_tokens: int = 1000,
):
    """Initialize Anthropic target.

    Args:
        model: Model name (claude-sonnet-4-20250514, etc.)
        api_key: API key (or uses ANTHROPIC_API_KEY env var)
        max_tokens: Maximum tokens in response
    """
    self.model = model
    self.api_key = api_key or os.environ.get("ANTHROPIC_API_KEY")
    self.max_tokens = max_tokens

    if not self.api_key:
        raise ValueError("Anthropic API key required")

query(prompt: str) -> TargetResponse async

Query Anthropic API.

Source code in src/rotalabs_redqueen/llm/targets.py 
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
async def query(self, prompt: str) -> TargetResponse:
    """Query Anthropic API."""
    import time

    import httpx

    start = time.time()

    async with httpx.AsyncClient() as client:
        try:
            response = await client.post(
                "https://api.anthropic.com/v1/messages",
                headers={
                    "x-api-key": self.api_key,
                    "anthropic-version": "2023-06-01",
                    "Content-Type": "application/json",
                },
                json={
                    "model": self.model,
                    "messages": [{"role": "user", "content": prompt}],
                    "max_tokens": self.max_tokens,
                },
                timeout=60.0,
            )

            if response.status_code == 429:
                raise RateLimitError("Anthropic rate limit exceeded")
            response.raise_for_status()

            data = response.json()
            latency = (time.time() - start) * 1000

            return TargetResponse(
                content=data["content"][0]["text"],
                model=self.model,
                tokens_used=data.get("usage", {}).get("input_tokens", 0)
                + data.get("usage", {}).get("output_tokens", 0),
                latency_ms=latency,
            )
        except httpx.HTTPStatusError as e:
            raise TargetError(f"Anthropic API error: {e}") from e
        except httpx.RequestError as e:
            raise NetworkError(f"Network error: {e}") from e

OllamaTarget

Bases: LLMTarget

Local Ollama target.

Source code in src/rotalabs_redqueen/llm/targets.py 
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
class OllamaTarget(LLMTarget):
    """Local Ollama target."""

    def __init__(
        self,
        model: str = "llama2",
        base_url: str = "http://localhost:11434",
    ):
        """Initialize Ollama target.

        Args:
            model: Model name
            base_url: Ollama API URL
        """
        self.model = model
        self.base_url = base_url

    @property
    def name(self) -> str:
        return f"ollama:{self.model}"

    async def query(self, prompt: str) -> TargetResponse:
        """Query Ollama API."""
        import time

        import httpx

        start = time.time()

        async with httpx.AsyncClient() as client:
            try:
                response = await client.post(
                    f"{self.base_url}/api/generate",
                    json={
                        "model": self.model,
                        "prompt": prompt,
                        "stream": False,
                    },
                    timeout=120.0,
                )
                response.raise_for_status()

                data = response.json()
                latency = (time.time() - start) * 1000

                return TargetResponse(
                    content=data["response"],
                    model=self.model,
                    tokens_used=data.get("eval_count", 0),
                    latency_ms=latency,
                )
            except httpx.HTTPStatusError as e:
                raise TargetError(f"Ollama API error: {e}") from e
            except httpx.RequestError as e:
                raise NetworkError(f"Network error: {e}") from e

__init__(model: str = 'llama2', base_url: str = 'http://localhost:11434')

Initialize Ollama target.

Parameters:

Name Type Description Default
model str

Model name

 'llama2'
base_url str

Ollama API URL

 'http://localhost:11434'
Source code in src/rotalabs_redqueen/llm/targets.py 
213
214
215
216
217
218
219
220
221
222
223
224
225
def __init__(
    self,
    model: str = "llama2",
    base_url: str = "http://localhost:11434",
):
    """Initialize Ollama target.

    Args:
        model: Model name
        base_url: Ollama API URL
    """
    self.model = model
    self.base_url = base_url

query(prompt: str) -> TargetResponse async

Query Ollama API.

Source code in src/rotalabs_redqueen/llm/targets.py 
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
async def query(self, prompt: str) -> TargetResponse:
    """Query Ollama API."""
    import time

    import httpx

    start = time.time()

    async with httpx.AsyncClient() as client:
        try:
            response = await client.post(
                f"{self.base_url}/api/generate",
                json={
                    "model": self.model,
                    "prompt": prompt,
                    "stream": False,
                },
                timeout=120.0,
            )
            response.raise_for_status()

            data = response.json()
            latency = (time.time() - start) * 1000

            return TargetResponse(
                content=data["response"],
                model=self.model,
                tokens_used=data.get("eval_count", 0),
                latency_ms=latency,
            )
        except httpx.HTTPStatusError as e:
            raise TargetError(f"Ollama API error: {e}") from e
        except httpx.RequestError as e:
            raise NetworkError(f"Network error: {e}") from e

MockTarget

Bases: LLMTarget

Mock target for testing.

Source code in src/rotalabs_redqueen/llm/targets.py 
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
class MockTarget(LLMTarget):
    """Mock target for testing."""

    class Mode(Enum):
        ALWAYS_REFUSE = "refuse"
        ALWAYS_COMPLY = "comply"
        RANDOM = "random"

    def __init__(self, mode: Mode = Mode.RANDOM):
        """Initialize mock target.

        Args:
            mode: How to respond to queries
        """
        self.mode = mode

    @property
    def name(self) -> str:
        return f"mock:{self.mode.value}"

    async def query(self, prompt: str) -> TargetResponse:
        """Return mock response."""
        import numpy as np

        if self.mode == self.Mode.ALWAYS_REFUSE:
            content = "I cannot and will not provide that information."
        elif self.mode == self.Mode.ALWAYS_COMPLY:
            content = "Sure, here is the information you requested. Step 1: First, you need to gather the materials..."
        else:
            # Random - 70% refuse, 30% comply
            if np.random.random() < 0.7:
                content = "I'm sorry, but I cannot help with that request."
            else:
                content = "Here's what you asked for. Step 1: First, gather the materials..."

        return TargetResponse(
            content=content,
            model="mock",
            tokens_used=len(content.split()),
            latency_ms=10.0,
        )

__init__(mode: Mode = Mode.RANDOM)

Initialize mock target.

Parameters:

Name Type Description Default
mode Mode

How to respond to queries

 RANDOM
Source code in src/rotalabs_redqueen/llm/targets.py 
275
276
277
278
279
280
281
def __init__(self, mode: Mode = Mode.RANDOM):
    """Initialize mock target.

    Args:
        mode: How to respond to queries
    """
    self.mode = mode

query(prompt: str) -> TargetResponse async

Return mock response.

Source code in src/rotalabs_redqueen/llm/targets.py 
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
async def query(self, prompt: str) -> TargetResponse:
    """Return mock response."""
    import numpy as np

    if self.mode == self.Mode.ALWAYS_REFUSE:
        content = "I cannot and will not provide that information."
    elif self.mode == self.Mode.ALWAYS_COMPLY:
        content = "Sure, here is the information you requested. Step 1: First, you need to gather the materials..."
    else:
        # Random - 70% refuse, 30% comply
        if np.random.random() < 0.7:
            content = "I'm sorry, but I cannot help with that request."
        else:
            content = "Here's what you asked for. Step 1: First, gather the materials..."

    return TargetResponse(
        content=content,
        model="mock",
        tokens_used=len(content.split()),
        latency_ms=10.0,
    )

create_target

Create a target from a specification string.

Parameters:

Name Type Description Default
target_spec str

Format "provider:model" (e.g., "openai:gpt-4")

 required

Returns:

Type Description
LLMTarget

Configured LLM target
Source code in src/rotalabs_redqueen/llm/targets.py 
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
def create_target(target_spec: str) -> LLMTarget:
    """Create a target from a specification string.

    Args:
        target_spec: Format "provider:model" (e.g., "openai:gpt-4")

    Returns:
        Configured LLM target
    """
    parts = target_spec.split(":", 1)
    provider = parts[0].lower()
    model = parts[1] if len(parts) > 1 else None

    if provider == "openai":
        return OpenAITarget(model=model or "gpt-4")
    elif provider == "anthropic":
        return AnthropicTarget(model=model or "claude-sonnet-4-20250514")
    elif provider == "ollama":
        return OllamaTarget(model=model or "llama2")
    elif provider == "mock":
        mode_map = {
            "refuse": MockTarget.Mode.ALWAYS_REFUSE,
            "comply": MockTarget.Mode.ALWAYS_COMPLY,
            "random": MockTarget.Mode.RANDOM,
        }
        mode = mode_map.get(model or "random", MockTarget.Mode.RANDOM)
        return MockTarget(mode=mode)
    else:
        raise ValueError(f"Unknown provider: {provider}")

TargetResponse

Response from LLM target.

Source code in src/rotalabs_redqueen/llm/targets.py 
32
33
34
35
36
37
38
39
@dataclass
class TargetResponse:
    """Response from LLM target."""

    content: str
    model: str
    tokens_used: int = 0
    latency_ms: float = 0.0

Errors

TargetError

Bases: Exception

Error from LLM target.

Source code in src/rotalabs_redqueen/llm/targets.py 
14
15
16
17
class TargetError(Exception):
    """Error from LLM target."""

    pass

RateLimitError

Bases: TargetError

Rate limit exceeded.

Source code in src/rotalabs_redqueen/llm/targets.py 
20
21
22
23
class RateLimitError(TargetError):
    """Rate limit exceeded."""

    pass

NetworkError

Bases: TargetError

Network connectivity error.

Source code in src/rotalabs_redqueen/llm/targets.py 
26
27
28
29
class NetworkError(TargetError):
    """Network connectivity error."""

    pass