Skip to navigation
LLM Proxy APIRed Team

Start Red Team Evaluation

View as Markdown

[BETA] This endpoint is not GA or Production ready and is subject to changes at any time. Breaking changes may occur.

Start a new red team client workflow. Auto-triggers planning phase. Client should then poll /next-action.

Authentication

AuthorizationBearer

Bearer authentication of the form Bearer <token>, where token is your auth token.

Request

This endpoint expects an object.
namestringRequired
Name for this evaluation
target_modelstringRequired
Target model identifier
objective_idslist of stringsOptional
Objective IDs to evaluate
execution_strategy_typestringOptionalDefaults to single
Execution strategy type
n_random_techniquesintegerOptionalDefaults to 1
Number of random techniques to use
max_parallel_techniquesintegerOptional
Maximum parallel techniques
max_turnsintegerOptionalDefaults to 5
Maximum conversation turns
attacker_max_generation_attemptsintegerOptionalDefaults to 3
Maximum generation attempts for attacker
refusal_judge_modelstringOptional
Model for refusal judging
objective_judge_modelstringOptional
Model for objective judging
evaluation_report_modelstringOptional
Model for evaluation report
attacker_modelstringOptional
Model for attacker
hl_project_idstringOptional
HiddenLayer project ID
prompt_set_idstringOptionalDefaults to
Prompt set ID for static prompt evaluation
sessions_per_techniqueintegerOptional
Number of sessions per technique
target_system_promptstringOptionalDefaults to
System prompt for the target

Response

Workflow started successfully
workflow_idstring
Workflow identifier
run_idstring
Run identifier

Errors

400
Bad Request Error
401
Unauthorized Error
403
Forbidden Error
422
Unprocessable Entity Error