hackagent 0.3.1__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- hackagent/__init__.py +12 -0
- hackagent/agent.py +214 -0
- hackagent/api/__init__.py +1 -0
- hackagent/api/agent/__init__.py +1 -0
- hackagent/api/agent/agent_create.py +347 -0
- hackagent/api/agent/agent_destroy.py +140 -0
- hackagent/api/agent/agent_list.py +242 -0
- hackagent/api/agent/agent_partial_update.py +361 -0
- hackagent/api/agent/agent_retrieve.py +235 -0
- hackagent/api/agent/agent_update.py +361 -0
- hackagent/api/apilogs/__init__.py +1 -0
- hackagent/api/apilogs/apilogs_list.py +170 -0
- hackagent/api/apilogs/apilogs_retrieve.py +162 -0
- hackagent/api/attack/__init__.py +1 -0
- hackagent/api/attack/attack_create.py +275 -0
- hackagent/api/attack/attack_destroy.py +146 -0
- hackagent/api/attack/attack_list.py +254 -0
- hackagent/api/attack/attack_partial_update.py +289 -0
- hackagent/api/attack/attack_retrieve.py +247 -0
- hackagent/api/attack/attack_update.py +289 -0
- hackagent/api/checkout/__init__.py +1 -0
- hackagent/api/checkout/checkout_create.py +225 -0
- hackagent/api/generate/__init__.py +1 -0
- hackagent/api/generate/generate_create.py +253 -0
- hackagent/api/judge/__init__.py +1 -0
- hackagent/api/judge/judge_create.py +253 -0
- hackagent/api/key/__init__.py +1 -0
- hackagent/api/key/key_create.py +179 -0
- hackagent/api/key/key_destroy.py +103 -0
- hackagent/api/key/key_list.py +170 -0
- hackagent/api/key/key_retrieve.py +162 -0
- hackagent/api/organization/__init__.py +1 -0
- hackagent/api/organization/organization_create.py +208 -0
- hackagent/api/organization/organization_destroy.py +104 -0
- hackagent/api/organization/organization_list.py +170 -0
- hackagent/api/organization/organization_me_retrieve.py +126 -0
- hackagent/api/organization/organization_partial_update.py +222 -0
- hackagent/api/organization/organization_retrieve.py +163 -0
- hackagent/api/organization/organization_update.py +222 -0
- hackagent/api/prompt/__init__.py +1 -0
- hackagent/api/prompt/prompt_create.py +171 -0
- hackagent/api/prompt/prompt_destroy.py +104 -0
- hackagent/api/prompt/prompt_list.py +185 -0
- hackagent/api/prompt/prompt_partial_update.py +185 -0
- hackagent/api/prompt/prompt_retrieve.py +163 -0
- hackagent/api/prompt/prompt_update.py +185 -0
- hackagent/api/result/__init__.py +1 -0
- hackagent/api/result/result_create.py +175 -0
- hackagent/api/result/result_destroy.py +106 -0
- hackagent/api/result/result_list.py +249 -0
- hackagent/api/result/result_partial_update.py +193 -0
- hackagent/api/result/result_retrieve.py +167 -0
- hackagent/api/result/result_trace_create.py +177 -0
- hackagent/api/result/result_update.py +189 -0
- hackagent/api/run/__init__.py +1 -0
- hackagent/api/run/run_create.py +187 -0
- hackagent/api/run/run_destroy.py +112 -0
- hackagent/api/run/run_list.py +291 -0
- hackagent/api/run/run_partial_update.py +201 -0
- hackagent/api/run/run_result_create.py +177 -0
- hackagent/api/run/run_retrieve.py +179 -0
- hackagent/api/run/run_run_tests_create.py +187 -0
- hackagent/api/run/run_update.py +201 -0
- hackagent/api/user/__init__.py +1 -0
- hackagent/api/user/user_create.py +212 -0
- hackagent/api/user/user_destroy.py +106 -0
- hackagent/api/user/user_list.py +174 -0
- hackagent/api/user/user_me_retrieve.py +126 -0
- hackagent/api/user/user_me_update.py +196 -0
- hackagent/api/user/user_partial_update.py +226 -0
- hackagent/api/user/user_retrieve.py +167 -0
- hackagent/api/user/user_update.py +226 -0
- hackagent/attacks/AdvPrefix/__init__.py +41 -0
- hackagent/attacks/AdvPrefix/completions.py +416 -0
- hackagent/attacks/AdvPrefix/config.py +259 -0
- hackagent/attacks/AdvPrefix/evaluation.py +745 -0
- hackagent/attacks/AdvPrefix/evaluators.py +564 -0
- hackagent/attacks/AdvPrefix/generate.py +711 -0
- hackagent/attacks/AdvPrefix/utils.py +307 -0
- hackagent/attacks/__init__.py +35 -0
- hackagent/attacks/advprefix.py +507 -0
- hackagent/attacks/base.py +106 -0
- hackagent/attacks/strategies.py +906 -0
- hackagent/cli/__init__.py +19 -0
- hackagent/cli/commands/__init__.py +20 -0
- hackagent/cli/commands/agent.py +100 -0
- hackagent/cli/commands/attack.py +417 -0
- hackagent/cli/commands/config.py +301 -0
- hackagent/cli/commands/results.py +327 -0
- hackagent/cli/config.py +249 -0
- hackagent/cli/main.py +515 -0
- hackagent/cli/tui/__init__.py +31 -0
- hackagent/cli/tui/actions_logger.py +200 -0
- hackagent/cli/tui/app.py +288 -0
- hackagent/cli/tui/base.py +137 -0
- hackagent/cli/tui/logger.py +318 -0
- hackagent/cli/tui/views/__init__.py +33 -0
- hackagent/cli/tui/views/agents.py +488 -0
- hackagent/cli/tui/views/attacks.py +624 -0
- hackagent/cli/tui/views/config.py +244 -0
- hackagent/cli/tui/views/dashboard.py +307 -0
- hackagent/cli/tui/views/results.py +1210 -0
- hackagent/cli/tui/widgets/__init__.py +24 -0
- hackagent/cli/tui/widgets/actions.py +346 -0
- hackagent/cli/tui/widgets/logs.py +435 -0
- hackagent/cli/utils.py +276 -0
- hackagent/client.py +286 -0
- hackagent/errors.py +37 -0
- hackagent/logger.py +83 -0
- hackagent/models/__init__.py +109 -0
- hackagent/models/agent.py +223 -0
- hackagent/models/agent_request.py +129 -0
- hackagent/models/api_token_log.py +184 -0
- hackagent/models/attack.py +154 -0
- hackagent/models/attack_request.py +82 -0
- hackagent/models/checkout_session_request_request.py +76 -0
- hackagent/models/checkout_session_response.py +59 -0
- hackagent/models/choice.py +81 -0
- hackagent/models/choice_message.py +67 -0
- hackagent/models/evaluation_status_enum.py +14 -0
- hackagent/models/generate_error_response.py +59 -0
- hackagent/models/generate_request_request.py +212 -0
- hackagent/models/generate_success_response.py +115 -0
- hackagent/models/generic_error_response.py +70 -0
- hackagent/models/message_request.py +67 -0
- hackagent/models/organization.py +102 -0
- hackagent/models/organization_minimal.py +68 -0
- hackagent/models/organization_request.py +71 -0
- hackagent/models/paginated_agent_list.py +123 -0
- hackagent/models/paginated_api_token_log_list.py +123 -0
- hackagent/models/paginated_attack_list.py +123 -0
- hackagent/models/paginated_organization_list.py +123 -0
- hackagent/models/paginated_prompt_list.py +123 -0
- hackagent/models/paginated_result_list.py +123 -0
- hackagent/models/paginated_run_list.py +123 -0
- hackagent/models/paginated_user_api_key_list.py +123 -0
- hackagent/models/paginated_user_profile_list.py +123 -0
- hackagent/models/patched_agent_request.py +128 -0
- hackagent/models/patched_attack_request.py +92 -0
- hackagent/models/patched_organization_request.py +71 -0
- hackagent/models/patched_prompt_request.py +125 -0
- hackagent/models/patched_result_request.py +237 -0
- hackagent/models/patched_run_request.py +138 -0
- hackagent/models/patched_user_profile_request.py +99 -0
- hackagent/models/prompt.py +220 -0
- hackagent/models/prompt_request.py +126 -0
- hackagent/models/result.py +294 -0
- hackagent/models/result_list_evaluation_status.py +14 -0
- hackagent/models/result_request.py +232 -0
- hackagent/models/run.py +233 -0
- hackagent/models/run_list_status.py +12 -0
- hackagent/models/run_request.py +133 -0
- hackagent/models/status_enum.py +12 -0
- hackagent/models/step_type_enum.py +14 -0
- hackagent/models/trace.py +121 -0
- hackagent/models/trace_request.py +94 -0
- hackagent/models/usage.py +75 -0
- hackagent/models/user_api_key.py +201 -0
- hackagent/models/user_api_key_request.py +73 -0
- hackagent/models/user_profile.py +135 -0
- hackagent/models/user_profile_minimal.py +76 -0
- hackagent/models/user_profile_request.py +99 -0
- hackagent/router/__init__.py +25 -0
- hackagent/router/adapters/__init__.py +20 -0
- hackagent/router/adapters/base.py +63 -0
- hackagent/router/adapters/google_adk.py +671 -0
- hackagent/router/adapters/litellm_adapter.py +524 -0
- hackagent/router/adapters/openai_adapter.py +426 -0
- hackagent/router/router.py +969 -0
- hackagent/router/types.py +54 -0
- hackagent/tracking/__init__.py +42 -0
- hackagent/tracking/context.py +163 -0
- hackagent/tracking/decorators.py +299 -0
- hackagent/tracking/tracker.py +441 -0
- hackagent/types.py +54 -0
- hackagent/utils.py +194 -0
- hackagent/vulnerabilities/__init__.py +13 -0
- hackagent/vulnerabilities/prompts.py +81 -0
- hackagent-0.3.1.dist-info/METADATA +122 -0
- hackagent-0.3.1.dist-info/RECORD +183 -0
- hackagent-0.3.1.dist-info/WHEEL +4 -0
- hackagent-0.3.1.dist-info/entry_points.txt +2 -0
- hackagent-0.3.1.dist-info/licenses/LICENSE +202 -0
|
@@ -0,0 +1,220 @@
|
|
|
1
|
+
import datetime
|
|
2
|
+
from collections.abc import Mapping
|
|
3
|
+
from typing import TYPE_CHECKING, Any, TypeVar, Union, cast
|
|
4
|
+
from uuid import UUID
|
|
5
|
+
|
|
6
|
+
from attrs import define as _attrs_define
|
|
7
|
+
from attrs import field as _attrs_field
|
|
8
|
+
from dateutil.parser import isoparse
|
|
9
|
+
|
|
10
|
+
from ..types import UNSET, Unset
|
|
11
|
+
|
|
12
|
+
if TYPE_CHECKING:
|
|
13
|
+
from ..models.organization_minimal import OrganizationMinimal
|
|
14
|
+
from ..models.user_profile_minimal import UserProfileMinimal
|
|
15
|
+
|
|
16
|
+
|
|
17
|
+
T = TypeVar("T", bound="Prompt")
|
|
18
|
+
|
|
19
|
+
|
|
20
|
+
@_attrs_define
|
|
21
|
+
class Prompt:
|
|
22
|
+
"""Serializer for the Prompt model.
|
|
23
|
+
|
|
24
|
+
Attributes:
|
|
25
|
+
id (UUID):
|
|
26
|
+
name (str):
|
|
27
|
+
prompt_text (str):
|
|
28
|
+
organization (UUID):
|
|
29
|
+
organization_detail (OrganizationMinimal):
|
|
30
|
+
owner (Union[None, int]):
|
|
31
|
+
owner_detail (Union['UserProfileMinimal', None]):
|
|
32
|
+
created_at (datetime.datetime):
|
|
33
|
+
updated_at (datetime.datetime):
|
|
34
|
+
category (Union[Unset, str]): Primary category for grouping prompts (e.g., Evasion, Harmful Content).
|
|
35
|
+
tags (Union[Unset, Any]): Optional JSON list of tags for classification (e.g., ["PII", "Malware"])
|
|
36
|
+
evaluation_criteria (Union[Unset, str]): Description of how success/failure should be judged for this specific
|
|
37
|
+
prompt.
|
|
38
|
+
expected_tool_calls (Union[Unset, Any]): JSON list of expected tool calls, e.g., [{"tool_name": "...",
|
|
39
|
+
"tool_input": {...}}]
|
|
40
|
+
expected_output_pattern (Union[Unset, str]): Optional regex pattern to match against the final response.
|
|
41
|
+
reference_output (Union[Unset, str]): Optional ideal/reference final output text.
|
|
42
|
+
"""
|
|
43
|
+
|
|
44
|
+
id: UUID
|
|
45
|
+
name: str
|
|
46
|
+
prompt_text: str
|
|
47
|
+
organization: UUID
|
|
48
|
+
organization_detail: "OrganizationMinimal"
|
|
49
|
+
owner: Union[None, int]
|
|
50
|
+
owner_detail: Union["UserProfileMinimal", None]
|
|
51
|
+
created_at: datetime.datetime
|
|
52
|
+
updated_at: datetime.datetime
|
|
53
|
+
category: Union[Unset, str] = UNSET
|
|
54
|
+
tags: Union[Unset, Any] = UNSET
|
|
55
|
+
evaluation_criteria: Union[Unset, str] = UNSET
|
|
56
|
+
expected_tool_calls: Union[Unset, Any] = UNSET
|
|
57
|
+
expected_output_pattern: Union[Unset, str] = UNSET
|
|
58
|
+
reference_output: Union[Unset, str] = UNSET
|
|
59
|
+
additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
|
|
60
|
+
|
|
61
|
+
def to_dict(self) -> dict[str, Any]:
|
|
62
|
+
from ..models.user_profile_minimal import UserProfileMinimal
|
|
63
|
+
|
|
64
|
+
id = str(self.id)
|
|
65
|
+
|
|
66
|
+
name = self.name
|
|
67
|
+
|
|
68
|
+
prompt_text = self.prompt_text
|
|
69
|
+
|
|
70
|
+
organization = str(self.organization)
|
|
71
|
+
|
|
72
|
+
organization_detail = self.organization_detail.to_dict()
|
|
73
|
+
|
|
74
|
+
owner: Union[None, int]
|
|
75
|
+
owner = self.owner
|
|
76
|
+
|
|
77
|
+
owner_detail: Union[None, dict[str, Any]]
|
|
78
|
+
if isinstance(self.owner_detail, UserProfileMinimal):
|
|
79
|
+
owner_detail = self.owner_detail.to_dict()
|
|
80
|
+
else:
|
|
81
|
+
owner_detail = self.owner_detail
|
|
82
|
+
|
|
83
|
+
created_at = self.created_at.isoformat()
|
|
84
|
+
|
|
85
|
+
updated_at = self.updated_at.isoformat()
|
|
86
|
+
|
|
87
|
+
category = self.category
|
|
88
|
+
|
|
89
|
+
tags = self.tags
|
|
90
|
+
|
|
91
|
+
evaluation_criteria = self.evaluation_criteria
|
|
92
|
+
|
|
93
|
+
expected_tool_calls = self.expected_tool_calls
|
|
94
|
+
|
|
95
|
+
expected_output_pattern = self.expected_output_pattern
|
|
96
|
+
|
|
97
|
+
reference_output = self.reference_output
|
|
98
|
+
|
|
99
|
+
field_dict: dict[str, Any] = {}
|
|
100
|
+
field_dict.update(self.additional_properties)
|
|
101
|
+
field_dict.update(
|
|
102
|
+
{
|
|
103
|
+
"id": id,
|
|
104
|
+
"name": name,
|
|
105
|
+
"prompt_text": prompt_text,
|
|
106
|
+
"organization": organization,
|
|
107
|
+
"organization_detail": organization_detail,
|
|
108
|
+
"owner": owner,
|
|
109
|
+
"owner_detail": owner_detail,
|
|
110
|
+
"created_at": created_at,
|
|
111
|
+
"updated_at": updated_at,
|
|
112
|
+
}
|
|
113
|
+
)
|
|
114
|
+
if category is not UNSET:
|
|
115
|
+
field_dict["category"] = category
|
|
116
|
+
if tags is not UNSET:
|
|
117
|
+
field_dict["tags"] = tags
|
|
118
|
+
if evaluation_criteria is not UNSET:
|
|
119
|
+
field_dict["evaluation_criteria"] = evaluation_criteria
|
|
120
|
+
if expected_tool_calls is not UNSET:
|
|
121
|
+
field_dict["expected_tool_calls"] = expected_tool_calls
|
|
122
|
+
if expected_output_pattern is not UNSET:
|
|
123
|
+
field_dict["expected_output_pattern"] = expected_output_pattern
|
|
124
|
+
if reference_output is not UNSET:
|
|
125
|
+
field_dict["reference_output"] = reference_output
|
|
126
|
+
|
|
127
|
+
return field_dict
|
|
128
|
+
|
|
129
|
+
@classmethod
|
|
130
|
+
def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
|
|
131
|
+
from ..models.organization_minimal import OrganizationMinimal
|
|
132
|
+
from ..models.user_profile_minimal import UserProfileMinimal
|
|
133
|
+
|
|
134
|
+
d = dict(src_dict)
|
|
135
|
+
id = UUID(d.pop("id"))
|
|
136
|
+
|
|
137
|
+
name = d.pop("name")
|
|
138
|
+
|
|
139
|
+
prompt_text = d.pop("prompt_text")
|
|
140
|
+
|
|
141
|
+
organization = UUID(d.pop("organization"))
|
|
142
|
+
|
|
143
|
+
organization_detail = OrganizationMinimal.from_dict(
|
|
144
|
+
d.pop("organization_detail")
|
|
145
|
+
)
|
|
146
|
+
|
|
147
|
+
def _parse_owner(data: object) -> Union[None, int]:
|
|
148
|
+
if data is None:
|
|
149
|
+
return data
|
|
150
|
+
return cast(Union[None, int], data)
|
|
151
|
+
|
|
152
|
+
owner = _parse_owner(d.pop("owner"))
|
|
153
|
+
|
|
154
|
+
def _parse_owner_detail(data: object) -> Union["UserProfileMinimal", None]:
|
|
155
|
+
if data is None:
|
|
156
|
+
return data
|
|
157
|
+
try:
|
|
158
|
+
if not isinstance(data, dict):
|
|
159
|
+
raise TypeError()
|
|
160
|
+
owner_detail_type_1 = UserProfileMinimal.from_dict(data)
|
|
161
|
+
|
|
162
|
+
return owner_detail_type_1
|
|
163
|
+
except: # noqa: E722
|
|
164
|
+
pass
|
|
165
|
+
return cast(Union["UserProfileMinimal", None], data)
|
|
166
|
+
|
|
167
|
+
owner_detail = _parse_owner_detail(d.pop("owner_detail"))
|
|
168
|
+
|
|
169
|
+
created_at = isoparse(d.pop("created_at"))
|
|
170
|
+
|
|
171
|
+
updated_at = isoparse(d.pop("updated_at"))
|
|
172
|
+
|
|
173
|
+
category = d.pop("category", UNSET)
|
|
174
|
+
|
|
175
|
+
tags = d.pop("tags", UNSET)
|
|
176
|
+
|
|
177
|
+
evaluation_criteria = d.pop("evaluation_criteria", UNSET)
|
|
178
|
+
|
|
179
|
+
expected_tool_calls = d.pop("expected_tool_calls", UNSET)
|
|
180
|
+
|
|
181
|
+
expected_output_pattern = d.pop("expected_output_pattern", UNSET)
|
|
182
|
+
|
|
183
|
+
reference_output = d.pop("reference_output", UNSET)
|
|
184
|
+
|
|
185
|
+
prompt = cls(
|
|
186
|
+
id=id,
|
|
187
|
+
name=name,
|
|
188
|
+
prompt_text=prompt_text,
|
|
189
|
+
organization=organization,
|
|
190
|
+
organization_detail=organization_detail,
|
|
191
|
+
owner=owner,
|
|
192
|
+
owner_detail=owner_detail,
|
|
193
|
+
created_at=created_at,
|
|
194
|
+
updated_at=updated_at,
|
|
195
|
+
category=category,
|
|
196
|
+
tags=tags,
|
|
197
|
+
evaluation_criteria=evaluation_criteria,
|
|
198
|
+
expected_tool_calls=expected_tool_calls,
|
|
199
|
+
expected_output_pattern=expected_output_pattern,
|
|
200
|
+
reference_output=reference_output,
|
|
201
|
+
)
|
|
202
|
+
|
|
203
|
+
prompt.additional_properties = d
|
|
204
|
+
return prompt
|
|
205
|
+
|
|
206
|
+
@property
|
|
207
|
+
def additional_keys(self) -> list[str]:
|
|
208
|
+
return list(self.additional_properties.keys())
|
|
209
|
+
|
|
210
|
+
def __getitem__(self, key: str) -> Any:
|
|
211
|
+
return self.additional_properties[key]
|
|
212
|
+
|
|
213
|
+
def __setitem__(self, key: str, value: Any) -> None:
|
|
214
|
+
self.additional_properties[key] = value
|
|
215
|
+
|
|
216
|
+
def __delitem__(self, key: str) -> None:
|
|
217
|
+
del self.additional_properties[key]
|
|
218
|
+
|
|
219
|
+
def __contains__(self, key: str) -> bool:
|
|
220
|
+
return key in self.additional_properties
|
|
@@ -0,0 +1,126 @@
|
|
|
1
|
+
from collections.abc import Mapping
|
|
2
|
+
from typing import Any, TypeVar, Union
|
|
3
|
+
|
|
4
|
+
from attrs import define as _attrs_define
|
|
5
|
+
from attrs import field as _attrs_field
|
|
6
|
+
|
|
7
|
+
from ..types import UNSET, Unset
|
|
8
|
+
|
|
9
|
+
T = TypeVar("T", bound="PromptRequest")
|
|
10
|
+
|
|
11
|
+
|
|
12
|
+
@_attrs_define
|
|
13
|
+
class PromptRequest:
|
|
14
|
+
"""Serializer for the Prompt model.
|
|
15
|
+
|
|
16
|
+
Attributes:
|
|
17
|
+
name (str):
|
|
18
|
+
prompt_text (str):
|
|
19
|
+
category (Union[Unset, str]): Primary category for grouping prompts (e.g., Evasion, Harmful Content).
|
|
20
|
+
tags (Union[Unset, Any]): Optional JSON list of tags for classification (e.g., ["PII", "Malware"])
|
|
21
|
+
evaluation_criteria (Union[Unset, str]): Description of how success/failure should be judged for this specific
|
|
22
|
+
prompt.
|
|
23
|
+
expected_tool_calls (Union[Unset, Any]): JSON list of expected tool calls, e.g., [{"tool_name": "...",
|
|
24
|
+
"tool_input": {...}}]
|
|
25
|
+
expected_output_pattern (Union[Unset, str]): Optional regex pattern to match against the final response.
|
|
26
|
+
reference_output (Union[Unset, str]): Optional ideal/reference final output text.
|
|
27
|
+
"""
|
|
28
|
+
|
|
29
|
+
name: str
|
|
30
|
+
prompt_text: str
|
|
31
|
+
category: Union[Unset, str] = UNSET
|
|
32
|
+
tags: Union[Unset, Any] = UNSET
|
|
33
|
+
evaluation_criteria: Union[Unset, str] = UNSET
|
|
34
|
+
expected_tool_calls: Union[Unset, Any] = UNSET
|
|
35
|
+
expected_output_pattern: Union[Unset, str] = UNSET
|
|
36
|
+
reference_output: Union[Unset, str] = UNSET
|
|
37
|
+
additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
|
|
38
|
+
|
|
39
|
+
def to_dict(self) -> dict[str, Any]:
|
|
40
|
+
name = self.name
|
|
41
|
+
|
|
42
|
+
prompt_text = self.prompt_text
|
|
43
|
+
|
|
44
|
+
category = self.category
|
|
45
|
+
|
|
46
|
+
tags = self.tags
|
|
47
|
+
|
|
48
|
+
evaluation_criteria = self.evaluation_criteria
|
|
49
|
+
|
|
50
|
+
expected_tool_calls = self.expected_tool_calls
|
|
51
|
+
|
|
52
|
+
expected_output_pattern = self.expected_output_pattern
|
|
53
|
+
|
|
54
|
+
reference_output = self.reference_output
|
|
55
|
+
|
|
56
|
+
field_dict: dict[str, Any] = {}
|
|
57
|
+
field_dict.update(self.additional_properties)
|
|
58
|
+
field_dict.update(
|
|
59
|
+
{
|
|
60
|
+
"name": name,
|
|
61
|
+
"prompt_text": prompt_text,
|
|
62
|
+
}
|
|
63
|
+
)
|
|
64
|
+
if category is not UNSET:
|
|
65
|
+
field_dict["category"] = category
|
|
66
|
+
if tags is not UNSET:
|
|
67
|
+
field_dict["tags"] = tags
|
|
68
|
+
if evaluation_criteria is not UNSET:
|
|
69
|
+
field_dict["evaluation_criteria"] = evaluation_criteria
|
|
70
|
+
if expected_tool_calls is not UNSET:
|
|
71
|
+
field_dict["expected_tool_calls"] = expected_tool_calls
|
|
72
|
+
if expected_output_pattern is not UNSET:
|
|
73
|
+
field_dict["expected_output_pattern"] = expected_output_pattern
|
|
74
|
+
if reference_output is not UNSET:
|
|
75
|
+
field_dict["reference_output"] = reference_output
|
|
76
|
+
|
|
77
|
+
return field_dict
|
|
78
|
+
|
|
79
|
+
@classmethod
|
|
80
|
+
def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
|
|
81
|
+
d = dict(src_dict)
|
|
82
|
+
name = d.pop("name")
|
|
83
|
+
|
|
84
|
+
prompt_text = d.pop("prompt_text")
|
|
85
|
+
|
|
86
|
+
category = d.pop("category", UNSET)
|
|
87
|
+
|
|
88
|
+
tags = d.pop("tags", UNSET)
|
|
89
|
+
|
|
90
|
+
evaluation_criteria = d.pop("evaluation_criteria", UNSET)
|
|
91
|
+
|
|
92
|
+
expected_tool_calls = d.pop("expected_tool_calls", UNSET)
|
|
93
|
+
|
|
94
|
+
expected_output_pattern = d.pop("expected_output_pattern", UNSET)
|
|
95
|
+
|
|
96
|
+
reference_output = d.pop("reference_output", UNSET)
|
|
97
|
+
|
|
98
|
+
prompt_request = cls(
|
|
99
|
+
name=name,
|
|
100
|
+
prompt_text=prompt_text,
|
|
101
|
+
category=category,
|
|
102
|
+
tags=tags,
|
|
103
|
+
evaluation_criteria=evaluation_criteria,
|
|
104
|
+
expected_tool_calls=expected_tool_calls,
|
|
105
|
+
expected_output_pattern=expected_output_pattern,
|
|
106
|
+
reference_output=reference_output,
|
|
107
|
+
)
|
|
108
|
+
|
|
109
|
+
prompt_request.additional_properties = d
|
|
110
|
+
return prompt_request
|
|
111
|
+
|
|
112
|
+
@property
|
|
113
|
+
def additional_keys(self) -> list[str]:
|
|
114
|
+
return list(self.additional_properties.keys())
|
|
115
|
+
|
|
116
|
+
def __getitem__(self, key: str) -> Any:
|
|
117
|
+
return self.additional_properties[key]
|
|
118
|
+
|
|
119
|
+
def __setitem__(self, key: str, value: Any) -> None:
|
|
120
|
+
self.additional_properties[key] = value
|
|
121
|
+
|
|
122
|
+
def __delitem__(self, key: str) -> None:
|
|
123
|
+
del self.additional_properties[key]
|
|
124
|
+
|
|
125
|
+
def __contains__(self, key: str) -> bool:
|
|
126
|
+
return key in self.additional_properties
|
|
@@ -0,0 +1,294 @@
|
|
|
1
|
+
import datetime
|
|
2
|
+
from collections.abc import Mapping
|
|
3
|
+
from typing import TYPE_CHECKING, Any, TypeVar, Union, cast
|
|
4
|
+
from uuid import UUID
|
|
5
|
+
|
|
6
|
+
from attrs import define as _attrs_define
|
|
7
|
+
from attrs import field as _attrs_field
|
|
8
|
+
from dateutil.parser import isoparse
|
|
9
|
+
|
|
10
|
+
from ..models.evaluation_status_enum import EvaluationStatusEnum
|
|
11
|
+
from ..types import UNSET, Unset
|
|
12
|
+
|
|
13
|
+
if TYPE_CHECKING:
|
|
14
|
+
from ..models.trace import Trace
|
|
15
|
+
|
|
16
|
+
|
|
17
|
+
T = TypeVar("T", bound="Result")
|
|
18
|
+
|
|
19
|
+
|
|
20
|
+
@_attrs_define
|
|
21
|
+
class Result:
|
|
22
|
+
"""Serializer for the Result model, often nested in RunSerializer.
|
|
23
|
+
|
|
24
|
+
Attributes:
|
|
25
|
+
id (UUID):
|
|
26
|
+
run (UUID):
|
|
27
|
+
run_id (UUID):
|
|
28
|
+
prompt_name (Union[None, str]):
|
|
29
|
+
timestamp (datetime.datetime):
|
|
30
|
+
traces (list['Trace']):
|
|
31
|
+
prompt (Union[None, UUID, Unset]):
|
|
32
|
+
request_payload (Union[Unset, Any]): Payload sent to agent or relevant data for client-submitted results.
|
|
33
|
+
response_status_code (Union[None, Unset, int]):
|
|
34
|
+
response_headers (Union[Unset, Any]):
|
|
35
|
+
response_body (Union[None, Unset, str]):
|
|
36
|
+
latency_ms (Union[None, Unset, int]):
|
|
37
|
+
detected_tool_calls (Union[Unset, Any]):
|
|
38
|
+
evaluation_status (Union[Unset, EvaluationStatusEnum]): * `NOT_EVALUATED` - Not Evaluated
|
|
39
|
+
* `SUCCESSFUL_JAILBREAK` - Successful Jailbreak
|
|
40
|
+
* `FAILED_JAILBREAK` - Failed Jailbreak (Mitigated/Refused)
|
|
41
|
+
* `ERROR_AGENT_RESPONSE` - Error in Agent Response
|
|
42
|
+
* `ERROR_TEST_FRAMEWORK` - Error in Test Framework
|
|
43
|
+
* `PASSED_CRITERIA` - Passed Criteria (Not Jailbreak)
|
|
44
|
+
* `FAILED_CRITERIA` - Failed Criteria (Not Jailbreak)
|
|
45
|
+
evaluation_notes (Union[Unset, str]):
|
|
46
|
+
evaluation_metrics (Union[Unset, Any]):
|
|
47
|
+
agent_specific_data (Union[Unset, Any]):
|
|
48
|
+
"""
|
|
49
|
+
|
|
50
|
+
id: UUID
|
|
51
|
+
run: UUID
|
|
52
|
+
run_id: UUID
|
|
53
|
+
prompt_name: Union[None, str]
|
|
54
|
+
timestamp: datetime.datetime
|
|
55
|
+
traces: list["Trace"]
|
|
56
|
+
prompt: Union[None, UUID, Unset] = UNSET
|
|
57
|
+
request_payload: Union[Unset, Any] = UNSET
|
|
58
|
+
response_status_code: Union[None, Unset, int] = UNSET
|
|
59
|
+
response_headers: Union[Unset, Any] = UNSET
|
|
60
|
+
response_body: Union[None, Unset, str] = UNSET
|
|
61
|
+
latency_ms: Union[None, Unset, int] = UNSET
|
|
62
|
+
detected_tool_calls: Union[Unset, Any] = UNSET
|
|
63
|
+
evaluation_status: Union[Unset, EvaluationStatusEnum] = UNSET
|
|
64
|
+
evaluation_notes: Union[Unset, str] = UNSET
|
|
65
|
+
evaluation_metrics: Union[Unset, Any] = UNSET
|
|
66
|
+
agent_specific_data: Union[Unset, Any] = UNSET
|
|
67
|
+
additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
|
|
68
|
+
|
|
69
|
+
def to_dict(self) -> dict[str, Any]:
|
|
70
|
+
id = str(self.id)
|
|
71
|
+
|
|
72
|
+
run = str(self.run)
|
|
73
|
+
|
|
74
|
+
run_id = str(self.run_id)
|
|
75
|
+
|
|
76
|
+
prompt_name: Union[None, str]
|
|
77
|
+
prompt_name = self.prompt_name
|
|
78
|
+
|
|
79
|
+
timestamp = self.timestamp.isoformat()
|
|
80
|
+
|
|
81
|
+
traces = []
|
|
82
|
+
for traces_item_data in self.traces:
|
|
83
|
+
traces_item = traces_item_data.to_dict()
|
|
84
|
+
traces.append(traces_item)
|
|
85
|
+
|
|
86
|
+
prompt: Union[None, Unset, str]
|
|
87
|
+
if isinstance(self.prompt, Unset):
|
|
88
|
+
prompt = UNSET
|
|
89
|
+
elif isinstance(self.prompt, UUID):
|
|
90
|
+
prompt = str(self.prompt)
|
|
91
|
+
else:
|
|
92
|
+
prompt = self.prompt
|
|
93
|
+
|
|
94
|
+
request_payload = self.request_payload
|
|
95
|
+
|
|
96
|
+
response_status_code: Union[None, Unset, int]
|
|
97
|
+
if isinstance(self.response_status_code, Unset):
|
|
98
|
+
response_status_code = UNSET
|
|
99
|
+
else:
|
|
100
|
+
response_status_code = self.response_status_code
|
|
101
|
+
|
|
102
|
+
response_headers = self.response_headers
|
|
103
|
+
|
|
104
|
+
response_body: Union[None, Unset, str]
|
|
105
|
+
if isinstance(self.response_body, Unset):
|
|
106
|
+
response_body = UNSET
|
|
107
|
+
else:
|
|
108
|
+
response_body = self.response_body
|
|
109
|
+
|
|
110
|
+
latency_ms: Union[None, Unset, int]
|
|
111
|
+
if isinstance(self.latency_ms, Unset):
|
|
112
|
+
latency_ms = UNSET
|
|
113
|
+
else:
|
|
114
|
+
latency_ms = self.latency_ms
|
|
115
|
+
|
|
116
|
+
detected_tool_calls = self.detected_tool_calls
|
|
117
|
+
|
|
118
|
+
evaluation_status: Union[Unset, str] = UNSET
|
|
119
|
+
if not isinstance(self.evaluation_status, Unset):
|
|
120
|
+
evaluation_status = self.evaluation_status.value
|
|
121
|
+
|
|
122
|
+
evaluation_notes = self.evaluation_notes
|
|
123
|
+
|
|
124
|
+
evaluation_metrics = self.evaluation_metrics
|
|
125
|
+
|
|
126
|
+
agent_specific_data = self.agent_specific_data
|
|
127
|
+
|
|
128
|
+
field_dict: dict[str, Any] = {}
|
|
129
|
+
field_dict.update(self.additional_properties)
|
|
130
|
+
field_dict.update(
|
|
131
|
+
{
|
|
132
|
+
"id": id,
|
|
133
|
+
"run": run,
|
|
134
|
+
"run_id": run_id,
|
|
135
|
+
"prompt_name": prompt_name,
|
|
136
|
+
"timestamp": timestamp,
|
|
137
|
+
"traces": traces,
|
|
138
|
+
}
|
|
139
|
+
)
|
|
140
|
+
if prompt is not UNSET:
|
|
141
|
+
field_dict["prompt"] = prompt
|
|
142
|
+
if request_payload is not UNSET:
|
|
143
|
+
field_dict["request_payload"] = request_payload
|
|
144
|
+
if response_status_code is not UNSET:
|
|
145
|
+
field_dict["response_status_code"] = response_status_code
|
|
146
|
+
if response_headers is not UNSET:
|
|
147
|
+
field_dict["response_headers"] = response_headers
|
|
148
|
+
if response_body is not UNSET:
|
|
149
|
+
field_dict["response_body"] = response_body
|
|
150
|
+
if latency_ms is not UNSET:
|
|
151
|
+
field_dict["latency_ms"] = latency_ms
|
|
152
|
+
if detected_tool_calls is not UNSET:
|
|
153
|
+
field_dict["detected_tool_calls"] = detected_tool_calls
|
|
154
|
+
if evaluation_status is not UNSET:
|
|
155
|
+
field_dict["evaluation_status"] = evaluation_status
|
|
156
|
+
if evaluation_notes is not UNSET:
|
|
157
|
+
field_dict["evaluation_notes"] = evaluation_notes
|
|
158
|
+
if evaluation_metrics is not UNSET:
|
|
159
|
+
field_dict["evaluation_metrics"] = evaluation_metrics
|
|
160
|
+
if agent_specific_data is not UNSET:
|
|
161
|
+
field_dict["agent_specific_data"] = agent_specific_data
|
|
162
|
+
|
|
163
|
+
return field_dict
|
|
164
|
+
|
|
165
|
+
@classmethod
|
|
166
|
+
def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
|
|
167
|
+
from ..models.trace import Trace
|
|
168
|
+
|
|
169
|
+
d = dict(src_dict)
|
|
170
|
+
id = UUID(d.pop("id"))
|
|
171
|
+
|
|
172
|
+
run = UUID(d.pop("run"))
|
|
173
|
+
|
|
174
|
+
run_id = UUID(d.pop("run_id"))
|
|
175
|
+
|
|
176
|
+
def _parse_prompt_name(data: object) -> Union[None, str]:
|
|
177
|
+
if data is None:
|
|
178
|
+
return data
|
|
179
|
+
return cast(Union[None, str], data)
|
|
180
|
+
|
|
181
|
+
prompt_name = _parse_prompt_name(d.pop("prompt_name"))
|
|
182
|
+
|
|
183
|
+
timestamp = isoparse(d.pop("timestamp"))
|
|
184
|
+
|
|
185
|
+
traces = []
|
|
186
|
+
_traces = d.pop("traces")
|
|
187
|
+
for traces_item_data in _traces:
|
|
188
|
+
traces_item = Trace.from_dict(traces_item_data)
|
|
189
|
+
|
|
190
|
+
traces.append(traces_item)
|
|
191
|
+
|
|
192
|
+
def _parse_prompt(data: object) -> Union[None, UUID, Unset]:
|
|
193
|
+
if data is None:
|
|
194
|
+
return data
|
|
195
|
+
if isinstance(data, Unset):
|
|
196
|
+
return data
|
|
197
|
+
try:
|
|
198
|
+
if not isinstance(data, str):
|
|
199
|
+
raise TypeError()
|
|
200
|
+
prompt_type_0 = UUID(data)
|
|
201
|
+
|
|
202
|
+
return prompt_type_0
|
|
203
|
+
except: # noqa: E722
|
|
204
|
+
pass
|
|
205
|
+
return cast(Union[None, UUID, Unset], data)
|
|
206
|
+
|
|
207
|
+
prompt = _parse_prompt(d.pop("prompt", UNSET))
|
|
208
|
+
|
|
209
|
+
request_payload = d.pop("request_payload", UNSET)
|
|
210
|
+
|
|
211
|
+
def _parse_response_status_code(data: object) -> Union[None, Unset, int]:
|
|
212
|
+
if data is None:
|
|
213
|
+
return data
|
|
214
|
+
if isinstance(data, Unset):
|
|
215
|
+
return data
|
|
216
|
+
return cast(Union[None, Unset, int], data)
|
|
217
|
+
|
|
218
|
+
response_status_code = _parse_response_status_code(
|
|
219
|
+
d.pop("response_status_code", UNSET)
|
|
220
|
+
)
|
|
221
|
+
|
|
222
|
+
response_headers = d.pop("response_headers", UNSET)
|
|
223
|
+
|
|
224
|
+
def _parse_response_body(data: object) -> Union[None, Unset, str]:
|
|
225
|
+
if data is None:
|
|
226
|
+
return data
|
|
227
|
+
if isinstance(data, Unset):
|
|
228
|
+
return data
|
|
229
|
+
return cast(Union[None, Unset, str], data)
|
|
230
|
+
|
|
231
|
+
response_body = _parse_response_body(d.pop("response_body", UNSET))
|
|
232
|
+
|
|
233
|
+
def _parse_latency_ms(data: object) -> Union[None, Unset, int]:
|
|
234
|
+
if data is None:
|
|
235
|
+
return data
|
|
236
|
+
if isinstance(data, Unset):
|
|
237
|
+
return data
|
|
238
|
+
return cast(Union[None, Unset, int], data)
|
|
239
|
+
|
|
240
|
+
latency_ms = _parse_latency_ms(d.pop("latency_ms", UNSET))
|
|
241
|
+
|
|
242
|
+
detected_tool_calls = d.pop("detected_tool_calls", UNSET)
|
|
243
|
+
|
|
244
|
+
_evaluation_status = d.pop("evaluation_status", UNSET)
|
|
245
|
+
evaluation_status: Union[Unset, EvaluationStatusEnum]
|
|
246
|
+
if isinstance(_evaluation_status, Unset):
|
|
247
|
+
evaluation_status = UNSET
|
|
248
|
+
else:
|
|
249
|
+
evaluation_status = EvaluationStatusEnum(_evaluation_status)
|
|
250
|
+
|
|
251
|
+
evaluation_notes = d.pop("evaluation_notes", UNSET)
|
|
252
|
+
|
|
253
|
+
evaluation_metrics = d.pop("evaluation_metrics", UNSET)
|
|
254
|
+
|
|
255
|
+
agent_specific_data = d.pop("agent_specific_data", UNSET)
|
|
256
|
+
|
|
257
|
+
result = cls(
|
|
258
|
+
id=id,
|
|
259
|
+
run=run,
|
|
260
|
+
run_id=run_id,
|
|
261
|
+
prompt_name=prompt_name,
|
|
262
|
+
timestamp=timestamp,
|
|
263
|
+
traces=traces,
|
|
264
|
+
prompt=prompt,
|
|
265
|
+
request_payload=request_payload,
|
|
266
|
+
response_status_code=response_status_code,
|
|
267
|
+
response_headers=response_headers,
|
|
268
|
+
response_body=response_body,
|
|
269
|
+
latency_ms=latency_ms,
|
|
270
|
+
detected_tool_calls=detected_tool_calls,
|
|
271
|
+
evaluation_status=evaluation_status,
|
|
272
|
+
evaluation_notes=evaluation_notes,
|
|
273
|
+
evaluation_metrics=evaluation_metrics,
|
|
274
|
+
agent_specific_data=agent_specific_data,
|
|
275
|
+
)
|
|
276
|
+
|
|
277
|
+
result.additional_properties = d
|
|
278
|
+
return result
|
|
279
|
+
|
|
280
|
+
@property
|
|
281
|
+
def additional_keys(self) -> list[str]:
|
|
282
|
+
return list(self.additional_properties.keys())
|
|
283
|
+
|
|
284
|
+
def __getitem__(self, key: str) -> Any:
|
|
285
|
+
return self.additional_properties[key]
|
|
286
|
+
|
|
287
|
+
def __setitem__(self, key: str, value: Any) -> None:
|
|
288
|
+
self.additional_properties[key] = value
|
|
289
|
+
|
|
290
|
+
def __delitem__(self, key: str) -> None:
|
|
291
|
+
del self.additional_properties[key]
|
|
292
|
+
|
|
293
|
+
def __contains__(self, key: str) -> bool:
|
|
294
|
+
return key in self.additional_properties
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
from enum import Enum
|
|
2
|
+
|
|
3
|
+
|
|
4
|
+
class ResultListEvaluationStatus(str, Enum):
|
|
5
|
+
ERROR_AGENT_RESPONSE = "ERROR_AGENT_RESPONSE"
|
|
6
|
+
ERROR_TEST_FRAMEWORK = "ERROR_TEST_FRAMEWORK"
|
|
7
|
+
FAILED_CRITERIA = "FAILED_CRITERIA"
|
|
8
|
+
FAILED_JAILBREAK = "FAILED_JAILBREAK"
|
|
9
|
+
NOT_EVALUATED = "NOT_EVALUATED"
|
|
10
|
+
PASSED_CRITERIA = "PASSED_CRITERIA"
|
|
11
|
+
SUCCESSFUL_JAILBREAK = "SUCCESSFUL_JAILBREAK"
|
|
12
|
+
|
|
13
|
+
def __str__(self) -> str:
|
|
14
|
+
return str(self.value)
|