# Bulk Widget Operations Source: https://docs.cekura.ai/api-reference/dashboards/bulk-widgets post /dashboards/dashboards/{id}/bulk_widgets/ Bulk create, update, and delete widgets on a dashboard in a single request. # Create Dashboard Source: https://docs.cekura.ai/api-reference/dashboards/create-dashboard post /dashboards/dashboards/ Create a new analytics dashboard # Create Widget Source: https://docs.cekura.ai/api-reference/dashboards/create-widget post /dashboards/widgets/ Create a new dashboard widget # Delete Dashboard Source: https://docs.cekura.ai/api-reference/dashboards/delete-dashboard delete /dashboards/dashboards/{id}/ Delete an analytics dashboard # Delete Widget Source: https://docs.cekura.ai/api-reference/dashboards/delete-widget delete /dashboards/widgets/{id}/ Delete a dashboard widget # Duplicate Dashboard Source: https://docs.cekura.ai/api-reference/dashboards/duplicate-dashboard post /dashboards/dashboards/{id}/duplicate/ Duplicate a dashboard and all its widgets. # Get Dashboard Source: https://docs.cekura.ai/api-reference/dashboards/get-dashboard get /dashboards/dashboards/{id}/ Retrieve an analytics dashboard by ID # Get Widget Source: https://docs.cekura.ai/api-reference/dashboards/get-widget get /dashboards/widgets/{id}/ Retrieve a dashboard widget by ID # Get Widget Data Source: https://docs.cekura.ai/api-reference/dashboards/get-widget-data get /dashboards/widgets/{id}/data/ Get plot data for a widget. # List Dashboards Source: https://docs.cekura.ai/api-reference/dashboards/list-dashboards get /dashboards/dashboards/ List analytics dashboards # List Widgets Source: https://docs.cekura.ai/api-reference/dashboards/list-widgets get /dashboards/widgets/ List dashboard widgets # Preview Widget Data Source: https://docs.cekura.ai/api-reference/dashboards/preview-widget-data post /dashboards/widgets/preview_data/ Preview widget data without saving. # Update Dashboard Source: https://docs.cekura.ai/api-reference/dashboards/update-dashboard patch /dashboards/dashboards/{id}/ Update an analytics dashboard # Update Widget Source: https://docs.cekura.ai/api-reference/dashboards/update-widget patch /dashboards/widgets/{id}/ Update a dashboard widget # Bulk Widget Data Source: https://docs.cekura.ai/api-reference/dashboards/widget-bulk-data post /dashboards/widgets/bulk_data/ Fetch chart data for multiple widgets in one request — useful for refreshing a dashboard view in a single round-trip. # Add Metric Failure Mode Insight Category Example Calls to Labs Source: https://docs.cekura.ai/api-reference/observability/add-metric-failure-mode-insight-category-example-calls-to-labs post /observability/v1/metric-failure-mode-insights/categories/{category_slug}/fix-metric/ Sample the latest calls in one persistent failure category and add them to Labs for its metric, with `feedback` as the note and an optional corrected `expected_value` label (defaults to a pass). # Bulk Generate Metric Failure Mode Insights Source: https://docs.cekura.ai/api-reference/observability/bulk-generate-metric-failure-mode-insights post /observability/v1/metric-failure-mode-insights/bulk_generate/ Fan out audits to every eligible metric in the project. # Create Agent Improvement Session Source: https://docs.cekura.ai/api-reference/observability/create-agent-improvement-session post /observability/v1/agent-improvement-sessions/ Create + poll "Improve Agent" sessions launched from the Insights page. # Create Evaluators from Call Logs V3 Source: https://docs.cekura.ai/api-reference/observability/create-evaluators-from-call-logs-v3 post /observability/v3/call-logs/create_scenarios/ Turn up to 50 call logs (a list of ids, or a number of call logs) into evaluators and return them once every call log is processed. Call logs need a transcript. Usually takes 1–3 minutes. # Create Evaluators from Call Logs Source: https://docs.cekura.ai/api-reference/observability/create-scenarios-from-call-logs post /observability/v1/call-logs/create_scenarios/ Create scenarios from call logs in background using Celery tasks. > [!WARNING] > > Deprecated. Use [Create Evaluators from Call Logs (v3)](/api-reference/observability/create-evaluators-from-call-logs-v3) instead: it returns the result in one call, with no polling. # Get Progress For Evaluator Creation Source: https://docs.cekura.ai/api-reference/observability/create-scenarios-from-call-logs-progress get /observability/v1/call-logs/create_scenarios_progress/ Get the progress of background scenario creation by progress_id. > [!WARNING] > > Deprecated. Use [Create Evaluators from Call Logs (v3)](/api-reference/observability/create-evaluators-from-call-logs-v3) instead: it returns the result in one call, with no polling. # Delete Call Source: https://docs.cekura.ai/api-reference/observability/delete-call delete /observability/v1/call-logs/{id}/ Delete a call log # Delete Deep Research Insight Source: https://docs.cekura.ai/api-reference/observability/delete-deep-research-insight delete /observability/v1/deep-research-insights/{id}/ Delete a deep research insight # Delete Metric Failure Mode Insight Category Source: https://docs.cekura.ai/api-reference/observability/delete-metric-failure-mode-insight-category delete /observability/v1/metric-failure-mode-insights/categories/{category_slug}/delete/ Remove one persistent failure category and all of its stored call assignments. Future calls are classified only against the remaining categories. Scope with `project` and `metric_id`. ⚠ Cannot be undone. # Dismiss Deep Research Insight Mode Source: https://docs.cekura.ai/api-reference/observability/dismiss-deep-research-insight-mode post /observability/v1/deep-research-insights/dismiss_mode/ Record a thumbs-down on one failure mode. # Evaluate Metrics Source: https://docs.cekura.ai/api-reference/observability/evaluate-metrics post /observability/v1/call-logs/evaluate_metrics/ Evaluate metrics against call log transcripts # Generate Deep Research Insights Source: https://docs.cekura.ai/api-reference/observability/generate-deep-research-insights post /observability/v1/deep-research-insights/generate/ Queue a Deep Research audit; returns the new row in ``pending``. # Get Agent Improvement Session Source: https://docs.cekura.ai/api-reference/observability/get-agent-improvement-session get /observability/v1/agent-improvement-sessions/{id}/ Create + poll "Improve Agent" sessions launched from the Insights page. # Get Alert Review Detail Source: https://docs.cekura.ai/api-reference/observability/get-alert-review-detail get /observability/v1/alerts/{id}/review_detail/ Per-alert review data — config snapshot, timeline, paginated fires. # Get Alert Summarization Progress Source: https://docs.cekura.ai/api-reference/observability/get-alert-summarization-progress get /observability/v1/alerts/summarize_progress/ Alerts summarize progress > [!WARNING] > > Deprecated. Use [Summarize Alert (v3)](/api-reference/observability/summarize-alert-v3) instead: it returns the result in one call, with no polling. # Get Call Source: https://docs.cekura.ai/api-reference/observability/get-call get /observability/v2/call-logs/{id}/ Retrieve a call log by ID # Retrieve Execution Logs for a Call Log Source: https://docs.cekura.ai/api-reference/observability/get-call-log-logs get /observability/v1/call-logs/{id}/logs/ Retrieve the provider execution logs captured for a production call log. # Get Call Log Trace Source: https://docs.cekura.ai/api-reference/observability/get-call-log-trace get /observability/v1/call-logs/{id}/trace/ Proxy trace data from Grafana Tempo for a CallLog. # Get Deep Research Insight Source: https://docs.cekura.ai/api-reference/observability/get-deep-research-insight get /observability/v1/deep-research-insights/{id}/ Retrieve a deep research insight by ID # Get Metric Failure Mode Insights Available Dates Source: https://docs.cekura.ai/api-reference/observability/get-metric-failure-mode-insights-available-dates get /observability/v1/metric-failure-mode-insights/available_dates/ Distinct calendar dates (newest first) that have a generated # List Agent Improvement Sessions Source: https://docs.cekura.ai/api-reference/observability/list-agent-improvement-sessions get /observability/v1/agent-improvement-sessions/ Create + poll "Improve Agent" sessions launched from the Insights page. # List Alerts for Review Source: https://docs.cekura.ai/api-reference/observability/list-alerts-for-review get /observability/v1/alerts/review/ Aggregate fired-alert data for the review page. # List Calls Source: https://docs.cekura.ai/api-reference/observability/list-calls get /observability/v2/call-logs/ List call logs # List Deep Research Insights Source: https://docs.cekura.ai/api-reference/observability/list-deep-research-insights get /observability/v1/deep-research-insights/ List deep research insights # List Metric Failure Mode Insight Categories Source: https://docs.cekura.ai/api-reference/observability/list-metric-failure-mode-insight-categories get /observability/v1/metric-failure-mode-insights/categories/ Return the persistent failure categories for a project (optionally one metric), each with a call count and example call ids. Filter by `date_from` / `date_to` (ISO datetimes for an exact window, or bare dates for whole-day bounds) to count only assignments in that range; results are sorted by count descending. # List Metric Failure Mode Insight Category Calls Source: https://docs.cekura.ai/api-reference/observability/list-metric-failure-mode-insight-category-calls get /observability/v1/metric-failure-mode-insights/categories/{category_slug}/calls/ Return all call-log ids assigned to one persistent failure category, newest first. Use the same date bounds as the category list to expand the examples shown for the selected range. The accumulator retains at most 500 calls per metric. # List Metric Failure Rates Source: https://docs.cekura.ai/api-reference/observability/list-metric-failure-rates get /observability/v1/metric-failure-mode-insights/failure-rates/ Return the percentage of evaluated calls that failed each metric during the selected Insights date range. # Reevaluate Calls Source: https://docs.cekura.ai/api-reference/observability/reevaluate-calls post /observability/v1/call-logs/rerun_evaluation/ Re-run metric evaluation on a call log # Retrieve Deep Research Insights Pricing Source: https://docs.cekura.ai/api-reference/observability/retrieve-deep-research-insights-pricing get /observability/v1/deep-research-insights/pricing/ Flat credit cost of a Deep Research ``Generate`` press. # Send Calls Source: https://docs.cekura.ai/api-reference/observability/send-calls post /observability/v1/observe/ Primary ingestion endpoint for production call logs. Your agent (or its provider's post-call webhook) POSTs here with the transcript, recording URL, and metadata; Cekura stores it as a `CallLog`, schedules metric evaluation, and surfaces it in Observability. # Summarize Alert Source: https://docs.cekura.ai/api-reference/observability/summarize-alert post /observability/v1/alerts/{id}/summarize/ Kick off an LLM summary of the last N hours of fires for this alert. > [!WARNING] > > Deprecated. Use [Summarize Alert (v3)](/api-reference/observability/summarize-alert-v3) instead: it returns the result in one call, with no polling. # Summarize Alert V3 Source: https://docs.cekura.ai/api-reference/observability/summarize-alert-v3 post /observability/v3/alerts/{id}/summarize/ Summarize what the alert's fires over the last `hours` have in common and return the summary. Returns 400 when the alert did not fire in that window. Usually takes under 30 seconds. # Trigger Alert Summary Source: https://docs.cekura.ai/api-reference/observability/trigger-alert-summary get /observability/v1/alerts/trigger_summary/ Return fire counts per alert over a time window. # Update Call Log Source: https://docs.cekura.ai/api-reference/observability/update-call-log patch /observability/v2/call-logs/{id}/ Update editable source data for an existing production call log. # Vote on Call Log Metric Source: https://docs.cekura.ai/api-reference/observability/vote-call-log-metric post /observability/v1/call-logs/{id}/mark_metric_vote/ Submit a thumbs-up or thumbs-down vote on a metric evaluation for a call log # Get Billing Info Source: https://docs.cekura.ai/api-reference/organization/get-billing-info get /test_framework/billing/info/ Returns credits remaining, credits used, and subscription status for the organization. The organization is automatically determined from the API key. Only organization-scoped API keys are supported. The response separates recurring plan credits in `balance` from non-expiring purchased or complimentary credits in `extra_balance`. `total_balance` is the sum available to the organization. Usage normally consumes `balance` before `extra_balance` when the combined balance covers the charge. # Create Cronjob Source: https://docs.cekura.ai/api-reference/schedules/create-cronjob post /schedules/v1/cron-jobs/ Create a cron job that runs evaluators against your agent on a recurring schedule (daily regression tests, hourly smoke tests, weekly rollups, etc.). The job invokes the same logic as `run_scenarios` but on a timer. # Create Schedule Job Source: https://docs.cekura.ai/api-reference/schedules/create-schedule-job post /schedules/v1/cron-jobs/ Create a cron job that runs evaluators against your agent on a recurring schedule (daily regression tests, hourly smoke tests, weekly rollups, etc.). The job invokes the same logic as `run_scenarios` but on a timer. # Delete Schedule Job Source: https://docs.cekura.ai/api-reference/schedules/delete-schedule-job delete /schedules/v1/cron-jobs/{id}/ Delete a scheduled job # Get Schedule Job Source: https://docs.cekura.ai/api-reference/schedules/get-schedule-job get /schedules/v1/cron-jobs/{id}/ Retrieve a scheduled job by id # List Schedule Jobs Source: https://docs.cekura.ai/api-reference/schedules/list-schedule-jobs get /schedules/v1/cron-jobs/ List all scheduled jobs # Update Schedule Job Source: https://docs.cekura.ai/api-reference/schedules/update-schedule-job patch /schedules/v1/cron-jobs/{id}/ Partially update an existing scheduled job # End API Chat Source: https://docs.cekura.ai/api-reference/test_framework/api-chat-end post /test_framework/v1/runs/{id}/end-chat/ End an API chat run so it's scored right away. # Poll API Chat Turn Source: https://docs.cekura.ai/api-reference/test_framework/api-chat-poll-turn get /test_framework/v1/runs/{id}/chat/{turn_id}/ Poll a submitted turn for the simulated user's reply. # Submit API Chat Turn Source: https://docs.cekura.ai/api-reference/test_framework/api-chat-submit-turn post /test_framework/v1/runs/{id}/chat/ Submit your agent's message on an API chat run. The reply is generated in the background — poll the returned turn for it. # Ask About Result Failures Source: https://docs.cekura.ai/api-reference/test_framework/ask-about-failures post /test_framework/v1/results/{id}/ask_about_failures/ Ask a question about a result's failed runs. Returns 202 with a progress_id; poll Get Ask About Failures Progress for the evidence-grounded answer and the run IDs it cites. # Get Ask About Failures Progress Source: https://docs.cekura.ai/api-reference/test_framework/ask-about-failures-progress get /test_framework/v1/results/failure-analysis-progress/ Poll a queued failed-run analysis. When status is completed, the response includes the answer and the run IDs used as evidence. # Ask About Failures v3 Source: https://docs.cekura.ai/api-reference/test_framework/ask-about-failures-v3 post /test_framework/v3/results/{id}/ask_about_failures/ Ask a question about a result's failed runs and return an evidence-grounded answer together with the runs it cites. # Auto-fetch Agent from Provider Source: https://docs.cekura.ai/api-reference/test_framework/auto-fetch-agent post /test_framework/v2/aiagents/{id}/auto-fetch/ Re-fetch the full agent configuration from its connected provider (Retell, VAPI, ElevenLabs, Synthflow): system prompt, call settings, phone number, tools with mock data, dynamic variables and knowledge base files. Returns a `progress_id`; poll the import-progress endpoint for staged status. > [!WARNING] > > Deprecated. Use [Auto-fetch Agent from Provider (v3)](/api-reference/test_framework/re-import-agent-v3) instead: it returns the result in one call, with no polling. # Auto-fetch Agent V3 Source: https://docs.cekura.ai/api-reference/test_framework/auto-fetch-agent-v3 post /test_framework/v3/aiagents/{id}/auto-fetch/ Re-fetch the agent's configuration from its connected provider (Retell, VAPI, ElevenLabs, Synthflow, Bland) — system prompt, call settings, phone number, tools with mock data, dynamic variables and knowledge base — and return the import result. Updates the agent in place. Usually takes under 2 minutes. # Bulk Delete Results Source: https://docs.cekura.ai/api-reference/test_framework/bulk-delete-results post /test_framework/v1/results/bulk_delete/ ⚠ Irreversible. Soft-deletes each result (and its non-deleted runs) with the same side-effects as deleting a single result: outbound phone numbers are released when no queued runs remain, capacity-manager slots are freed, and active_sessions / pending_demand counters are decremented in aggregate. Maximum 100 result IDs per request. Results that are already deleted or that the caller cannot access are silently skipped — the response is 204 even when no results are deleted. # Bulk Delete Results External Source: https://docs.cekura.ai/api-reference/test_framework/bulk-delete-results-external post /test_framework/v1/results-external/bulk_delete/ ⚠ Irreversible. Soft-deletes each result (and its non-deleted runs) with the same side-effects as deleting a single result: outbound phone numbers are released when no queued runs remain, capacity-manager slots are freed, and active_sessions / pending_demand counters are decremented in aggregate. Maximum 100 result IDs per request. Results that are already deleted or that the caller cannot access are silently skipped — the response is 204 even when no results are deleted. # Bulk Delete Evaluators Source: https://docs.cekura.ai/api-reference/test_framework/bulk-delete-scenarios post /test_framework/v1/scenarios/delete_scenarios/ Bulk-delete multiple evaluators (test scenarios) in a single call. Cannot be undone from the UI — confirm the exact list with the user before calling. # Bulk Evaluate Results Metrics Source: https://docs.cekura.ai/api-reference/test_framework/bulk-evaluate-results-metrics post /test_framework/v1/results/bulk_evaluate_metrics/ Schedules re-evaluation of the given metrics against every completed/failed, non-deleted run of every selected result. Metrics are validated against each agent's metric scope; balance is checked for the full batch up front before any work is dispatched. Each affected result's status is flipped to RUNNING. Maximum 100 result IDs per request. # Bulk Evaluate Results Metrics External Source: https://docs.cekura.ai/api-reference/test_framework/bulk-evaluate-results-metrics-external post /test_framework/v1/results-external/bulk_evaluate_metrics/ Schedules re-evaluation of the given metrics against every completed/failed, non-deleted run of every selected result. Metrics are validated against each agent's metric scope; balance is checked for the full batch up front before any work is dispatched. Each affected result's status is flipped to RUNNING. Maximum 100 result IDs per request. # Bulk Update Evaluators Source: https://docs.cekura.ai/api-reference/test_framework/bulk-update-scenarios post /test_framework/v1/scenarios/adv_update/ Bulk-update many evaluators in a single call: attach/detach metrics or tools, switch personality, append instructions, move into a folder, or change phone number / max duration. # Correct External Scenario Condition Audio Source: https://docs.cekura.ai/api-reference/test_framework/correct-external-scenario-condition-audio patch /test_framework/v1/scenarios-external/{id}/condition-audio/{audio_id}/ Correct an audio clip's transcript and word timings # Correct Scenario Condition Audio Source: https://docs.cekura.ai/api-reference/test_framework/correct-scenario-condition-audio patch /test_framework/v1/scenarios/{id}/condition-audio/{audio_id}/ Correct an audio clip's transcript and word timings # Correct Scenario Condition Audio V2 Source: https://docs.cekura.ai/api-reference/test_framework/correct-scenario-condition-audio-v2 patch /test_framework/v2/scenarios/{id}/condition-audio/{audio_id}/ Correct an audio clip's transcript and word timings # Create Agent Source: https://docs.cekura.ai/api-reference/test_framework/create-agent post /test_framework/v2/aiagents/ Register an AI voice agent with Cekura. Provider credentials are nested under a provider block rather than flat fields. # Create Agent v3 Source: https://docs.cekura.ai/api-reference/test_framework/create-agent-v3 post /test_framework/v3/aiagents/ Create an agent and return it once it is ready. Takes the same fields as Create Agent (`POST /v2/aiagents/`). With `provider.configure_from_provider: true` the agent is first imported from its cloud provider, usually within 2 minutes; 409 when that assistant is already imported into the project. # Create Evaluator Source: https://docs.cekura.ai/api-reference/test_framework/create-evaluator post /test_framework/v1/scenarios/ Create a test scenario (evaluator) that simulates a caller interacting with your agent. # Create Evaluator from Transcript Source: https://docs.cekura.ai/api-reference/test_framework/create-evaluator-from-transcript-bg post /test_framework/v1/scenarios/create-from-transcript-bg/ Create a test scenario from a call transcript in the background > [!WARNING] > > Deprecated. Use [Create Evaluator from Transcript (v3)](/api-reference/test_framework/create-evaluator-from-transcript-v3) instead: it returns the result in one call, with no polling. # Get Create-from-Transcript Progress Source: https://docs.cekura.ai/api-reference/test_framework/create-evaluator-from-transcript-progress get /test_framework/v1/scenarios/create-from-transcript-progress/ Poll progress of a transcript-based evaluator creation > [!WARNING] > > Deprecated. Use [Create Evaluator from Transcript (v3)](/api-reference/test_framework/create-evaluator-from-transcript-v3) instead: it returns the result in one call, with no polling. # Create Evaluator from Transcript V3 Source: https://docs.cekura.ai/api-reference/test_framework/create-evaluator-from-transcript-v3 post /test_framework/v3/scenarios/create-from-transcript/ Create an evaluator from a call transcript and return it once it is ready. `transcript_json` is a list of `{role, content, start_time}` messages, where `role` is `Main Agent`, `Testing Agent`, `Function Call` or `Function Call Result`. Usually takes under 2 minutes. # Create External Personality Voice Clone Source: https://docs.cekura.ai/api-reference/test_framework/create-external-personality-voice-clone post /test_framework/v1/personalities-external/clone_voice/ Clone a voice from an uploaded audio sample and return its provider voice id. # Create Folder Source: https://docs.cekura.ai/api-reference/test_framework/create-folder post /test_framework/v1/scenarios/create_folder/ Create a new scenario folder # Create Scenario Improvement Session Source: https://docs.cekura.ai/api-reference/test_framework/create-improvement-session post /test_framework/v1/scenario-improvement-sessions/ Create a scenario-improvement session — a persistent container for a multi-step `scenarios_refine` improvement loop. # Create Metric Source: https://docs.cekura.ai/api-reference/test_framework/create-metric post /test_framework/v2/metrics/ Create a quality metric for evaluating agent conversations. # Create Metrics in Bulk Source: https://docs.cekura.ai/api-reference/test_framework/create-metrics-in-bulk post /test_framework/v1/metrics/bulk_create/ Create multiple metrics in a single request # Create Personality Source: https://docs.cekura.ai/api-reference/test_framework/create-personality post /test_framework/v1/personalities/ Create a new caller personality. Personalities define how the simulated caller speaks during runs (voice, accent, language, gender, instructions); they are then attached to evaluators. # Create Personality Voice Clone Source: https://docs.cekura.ai/api-reference/test_framework/create-personality-voice-clone post /test_framework/v1/personalities/clone_voice/ Clone a voice from an uploaded audio sample and return its provider voice id. # Create Test Profile Source: https://docs.cekura.ai/api-reference/test_framework/create-test-profile post /test_framework/v1/test-profiles/ Create a new test profile # Delete Agent Source: https://docs.cekura.ai/api-reference/test_framework/delete-agent delete /test_framework/v2/aiagents/{id}/ Delete an AI voice agent # Delete Evaluator Source: https://docs.cekura.ai/api-reference/test_framework/delete-evaluator delete /test_framework/v1/scenarios/{id}/ Delete a test scenario # Delete Folder Source: https://docs.cekura.ai/api-reference/test_framework/delete-folder post /test_framework/v1/scenarios/delete_folder/ Delete a scenario folder and its contents # Delete Metric Source: https://docs.cekura.ai/api-reference/test_framework/delete-metric delete /test_framework/v1/metrics/{id}/ Delete a quality metric # Delete Test Profile Source: https://docs.cekura.ai/api-reference/test_framework/delete-test-profile Delete /test_framework/v1/test-profiles/{id}/ Delete Test Profile # Duplicate Agent Source: https://docs.cekura.ai/api-reference/test_framework/duplicate-agent post /test_framework/v2/aiagents/{id}/duplicate/ Duplicate an AI voice agent with all its configuration # Fetch Scenarios JSON Schema Source: https://docs.cekura.ai/api-reference/test_framework/fetch-scenarios-json-schema get /test_framework/v1/scenarios/json_schema/ Fetch the JSON Schema for test-suite spec files # Fork Personality Source: https://docs.cekura.ai/api-reference/test_framework/fork-personality post /test_framework/v1/personalities/{id}/fork/ Copy an existing personality — including a globally available one — into one of your projects, where you can then change any of its settings. # Generate Clean Metric Description V3 Source: https://docs.cekura.ai/api-reference/test_framework/generate-clean-metric-description-v3 post /test_framework/v3/metrics/generate_clean_description/ Turn a rough metric description into a clean one suited to the agent, and return it. Usually takes under 30 seconds. # Generate Metric Evaluation Trigger Source: https://docs.cekura.ai/api-reference/test_framework/generate-evaluation-trigger post /test_framework/v1/metrics/generate_evaluation_trigger/ Use AI to derive an `evaluation_trigger` for a metric — the condition that decides when the metric should run on a transcript. Prefer this over guessing the trigger by hand. # Generate Evaluator Source: https://docs.cekura.ai/api-reference/test_framework/generate-evaluator-bg post /test_framework/v1/scenarios/generate-bg/ Generate test scenarios in the background > [!WARNING] > > Deprecated. Use [Generate Evaluator (v3)](/api-reference/test_framework/generate-evaluators-v3) instead: it returns the result in one call, with no polling. # Get Generate Evaluator Progress Source: https://docs.cekura.ai/api-reference/test_framework/generate-evaluator-poll-progress get /test_framework/v1/scenarios/generate-progress/ Check progress of AI scenario generation > [!WARNING] > > Deprecated. Use [Generate Evaluator (v3)](/api-reference/test_framework/generate-evaluators-v3) instead: it returns the result in one call, with no polling. # Generate Evaluator V3 Source: https://docs.cekura.ai/api-reference/test_framework/generate-evaluator-v3 post /test_framework/v3/scenarios/generate/ Generate evaluators for an agent and return them once generation finishes. Generation never pauses for clarification questions. Usually takes 1–4 minutes. # Generate Scenario Instructions with AI Source: https://docs.cekura.ai/api-reference/test_framework/generate-instructions post /test_framework/v1/scenarios/generate-instructions/ Generate a scenario's `instructions` text from agent context and natural-language input. Returns a `progress_id` — poll `instructions-progress/`. > [!WARNING] > > Deprecated. Use [Generate Scenario Instructions with AI (v3)](/api-reference/test_framework/generate-evaluator-instructions-v3) instead: it returns the result in one call, with no polling. # Generate Clean Metric Description Source: https://docs.cekura.ai/api-reference/test_framework/generate-metric-description post /test_framework/v1/metrics/generate_clean_description/ Use AI to clean up a metric `description` (the LLM-judge criterion) into a clear, well-formed prompt. Prefer this over hand-editing a metric's description via `metrics_partial_update`. # Generate Metrics Source: https://docs.cekura.ai/api-reference/test_framework/generate-metrics post /test_framework/v1/metrics/generate_metrics/ Generate metrics for an AI agent. Returns a progress_id that should be polled via the generate_metrics_progress endpoint. > [!WARNING] > > Deprecated. Use [Generate Metrics (v3)](/api-reference/test_framework/generate-suggested-metrics-v3) instead: it returns the result in one call, with no polling. > [!NOTE] > > This endpoint returns a `progress_id` — it does not create metrics directly. > > **Full workflow:** > > 1. Call this endpoint to start metric generation → receive a `progress_id` > 2. Poll [Generate Metrics Progress](/api-reference/test_framework/generate-metrics-progress) with the `progress_id` until `status` is `"completed"` > 3. Take the `metrics` array from the progress response and pass it to [Create Metrics in Bulk](/api-reference/test_framework/create-metrics-in-bulk) to save them # Get Generate Metrics Progress Source: https://docs.cekura.ai/api-reference/test_framework/generate-metrics-progress get /test_framework/v1/metrics/generate_metrics_progress/ Metrics generate metrics progress Deprecated: `POST /test_framework/v3/metrics/generate_metrics/` does this in one call and returns the result. > [!WARNING] > > Deprecated. Use [Generate Metrics (v3)](/api-reference/test_framework/generate-suggested-metrics-v3) instead: it returns the result in one call, with no polling. > [!NOTE] > > When `status` is `"completed"`, the response includes a `metrics` array with the generated metric definitions. These metrics are **not yet saved** — pass them to [Create Metrics in Bulk](/api-reference/test_framework/create-metrics-in-bulk) to persist them. # Generate Metrics V3 Source: https://docs.cekura.ai/api-reference/test_framework/generate-metrics-v3 post /test_framework/v3/metrics/generate_metrics/ Suggest metrics for an agent and return them. The metrics are not saved; pass them to Create Metrics in Bulk to keep them. The agent needs a description of at least 100 characters. Usually takes under a minute. # Generate Scenario Instructions V3 Source: https://docs.cekura.ai/api-reference/test_framework/generate-scenario-instructions-v3 post /test_framework/v3/scenarios/generate-instructions/ Write caller instructions and an expected outcome for a new evaluator of the agent, guided by `extra_instructions`, and return them. Nothing is saved. Usually takes under a minute. # Get Agent Source: https://docs.cekura.ai/api-reference/test_framework/get-agent get /test_framework/v2/aiagents/{id}/ Retrieve an AI voice agent by ID # Get Combined Results Report Source: https://docs.cekura.ai/api-reference/test_framework/get-combined-results-report get /test_framework/v2/results/reports/ Aggregate two or more test results into a single report with success rate, metric summary, critical categories, latency and failure reasons # Get Evaluator Source: https://docs.cekura.ai/api-reference/test_framework/get-evaluator get /test_framework/v1/scenarios/{id}/ Retrieve a test scenario by ID # Get Scenario Improvement Session Source: https://docs.cekura.ai/api-reference/test_framework/get-improvement-session get /test_framework/v1/scenario-improvement-sessions/{id}/ Retrieve a scenario-improvement session by ID, including its accumulated state and previous turns. # Get Metric Source: https://docs.cekura.ai/api-reference/test_framework/get-metric get /test_framework/v2/metrics/{id}/ Retrieve a specific metric by ID or slug. # Get Payment History Source: https://docs.cekura.ai/api-reference/test_framework/get-payment-history get /test_framework/billing/payment-history/ Returns a paginated list of all payments (auto-pay, one-time, subscription) made by the organization. # Get Result Source: https://docs.cekura.ai/api-reference/test_framework/get-result get /test_framework/v2/results/{id}/ Retrieve a test result by ID # Retrieve Execution Logs for a Simulation Run Source: https://docs.cekura.ai/api-reference/test_framework/get-run-logs get /test_framework/v1/runs/{id}/logs/ Retrieve the provider execution logs captured for a simulation run. # Get Run Trace Source: https://docs.cekura.ai/api-reference/test_framework/get-run-trace get /test_framework/v1/runs/{id}/trace/ Proxy trace data from Grafana Tempo for a Run. # Get Test Profile Source: https://docs.cekura.ai/api-reference/test_framework/get-test-profile get /test_framework/v1/test-profiles/{id}/ Retrieve a test profile by ID # Get Tool Source: https://docs.cekura.ai/api-reference/test_framework/get-tool get /test_framework/v1/tools/{id}/ Return a single mock tool with its input/output pairs and the agents using it. # Get Unit Usage Report Source: https://docs.cekura.ai/api-reference/test_framework/get-unit-usage-report get /test_framework/billing/unit-usage-report/ Returns an organization's usage per billable unit (voice and text simulations, monitored calls, metric evaluations, triage, insights, Labs and AI metric building) with the credits billed for each, one column per calendar month (UTC) plus a total. The range may span at most 12 months. Each row's `values` maps a month (`YYYY-MM`) to a number; `other_credits` covers any remaining billed item and refunds are a negative row, so credit rows sum to `total_credits`. With an organization-scoped API key the organization comes from the key; with a user session pass `organization_id` (organization admins only). Pass `format=csv` to download the same table as CSV. Results may lag usage by up to 30 minutes. # Improve Evaluators with AI Source: https://docs.cekura.ai/api-reference/test_framework/improve-evaluators post /test_framework/v1/scenarios/scenario-agent/ Improve evaluators using AI by providing natural language feedback. # Improve Scenario Instructions with AI Source: https://docs.cekura.ai/api-reference/test_framework/improve-instructions post /test_framework/v1/scenarios/improve_instructions/ Refine a scenario's `instructions` text using AI given the agent context and natural-language extra_instructions. Returns a `progress_id` — poll `instructions-progress/`. Lower-level than `scenario-agent/`; for multi-scenario or higher-level edits prefer `scenarios-agent-create`. > [!WARNING] > > Deprecated. Use [Improve Scenario Instructions with AI (v3)](/api-reference/test_framework/improve-evaluator-instructions-v3) instead: it returns the result in one call, with no polling. # Poll Improve/Generate Instructions Progress Source: https://docs.cekura.ai/api-reference/test_framework/improve-instructions-progress get /test_framework/v1/scenarios/instructions-progress/ Poll the progress of an asynchronous instruction generate/improve job started by `generate-instructions/` or `improve_instructions/`. > [!WARNING] > > Deprecated. Use [Generate Scenario Instructions with AI (v3)](/api-reference/test_framework/generate-evaluator-instructions-v3) or [Improve Scenario Instructions with AI (v3)](/api-reference/test_framework/improve-evaluator-instructions-v3) instead: they return the result in one call, with no polling. # Improve Scenario Instructions V3 Source: https://docs.cekura.ai/api-reference/test_framework/improve-scenario-instructions-v3 post /test_framework/v3/scenarios/improve_instructions/ Rewrite `current_instructions`, or write new ones, following `extra_instructions`, and return the result. Nothing is saved; apply it with the evaluator update endpoint. Usually takes under a minute. # List Agents Source: https://docs.cekura.ai/api-reference/test_framework/list-agents get /test_framework/v2/aiagents/ List AI voice agents # List Critical Metric Scenarios Source: https://docs.cekura.ai/api-reference/test_framework/list-critical-metric-scenarios get /test_framework/v1/critical-metric-scenarios/ List scenarios flagged by critical metrics # List Evaluators Source: https://docs.cekura.ai/api-reference/test_framework/list-evaluators get /test_framework/v1/scenarios/ List test scenarios # List Folders Source: https://docs.cekura.ai/api-reference/test_framework/list-folders get /test_framework/v1/scenarios/folders/ List scenario folders for a project # List Scenario Improvement Sessions Source: https://docs.cekura.ai/api-reference/test_framework/list-improvement-sessions get /test_framework/v1/scenario-improvement-sessions/ List scenario-improvement sessions. Each session groups a multi-step scenario-agent improvement workflow and can be resumed via `scenario-improvement-sessions/{id}/`. # List Inbound Phone Numbers Source: https://docs.cekura.ai/api-reference/test_framework/list-inbound-phone-numbers get /test_framework/v1/phone-numbers/inbound/ Phone numbers inbound # List Metrics Source: https://docs.cekura.ai/api-reference/test_framework/list-metrics get /test_framework/v2/metrics/ List quality metrics. Filter by `project_id` or `agent_id` to scope results. # List Personalities Source: https://docs.cekura.ai/api-reference/test_framework/list-personalities get /test_framework/v1/personalities/ List caller personalities # List Predefined Metrics Source: https://docs.cekura.ai/api-reference/test_framework/list-predefined-metrics get /test_framework/v1/predefined-metrics/ List predefined metrics # List Provider Assistants Source: https://docs.cekura.ai/api-reference/test_framework/list-provider-assistants get /test_framework/v2/aiagents/provider-assistants/ List every assistant/agent the project's saved provider API key can see on a cloud provider (vapi, retell, elevenlabs, synthflow, bland), flagging those already imported into the project. Import one with `POST /v2/aiagents/` + `provider.configure_from_provider`; when `provider.credentials.api_key` is omitted there, the same saved key is used. # List Results Source: https://docs.cekura.ai/api-reference/test_framework/list-results get /test_framework/v2/results/ List test results # List Runs with IDs Source: https://docs.cekura.ai/api-reference/test_framework/list-runs-with-ids get /test_framework/v2/runs/bulk/ Retrieve multiple runs in a single request # List Test Profiles Source: https://docs.cekura.ai/api-reference/test_framework/list-test-profiles get /test_framework/v1/test-profiles/ List test profiles # List Tools Source: https://docs.cekura.ai/api-reference/test_framework/list-tools get /test_framework/v1/tools/ Return the project's mock tools with their input/output pairs and the agents using each one. Scope with `project_id` for the whole catalogue, or with `agent_id` to get only the tools that agent could newly adopt (its own are excluded). `search` matches the tool name, with or without the `c_` prefix. Results are paginated, most-shared first. # Preview Metric Source: https://docs.cekura.ai/api-reference/test_framework/metric-preview post /test_framework/v1/metrics/preview/ Preview metric evaluation on call logs or runs before creating it. Supports up to 10 items. Returns progress_id to poll for results. > [!WARNING] > > Deprecated. Use [Preview Metric (v3)](/api-reference/test_framework/preview-metric-v3) instead: it returns the result in one call, with no polling. # Get Metric Preview Progress Source: https://docs.cekura.ai/api-reference/test_framework/metric-preview-progress get /test_framework/v1/metrics/preview-progress/ Poll the progress of metric preview. Returns current state with evaluation results. > [!WARNING] > > Deprecated. Use [Preview Metric (v3)](/api-reference/test_framework/preview-metric-v3) instead: it returns the result in one call, with no polling. # Auto-Improve Metric Source: https://docs.cekura.ai/api-reference/test_framework/metric-review-process-feedbacks post /test_framework/metric-reviews/process_feedbacks/ Improve a metric using feedback from test sets. Returns improved metric configuration. Use the Update Metric endpoint to save changes. # Get Auto-Improve Progress Source: https://docs.cekura.ai/api-reference/test_framework/metric-review-process-feedbacks-progress get /test_framework/metric-reviews/process_feedbacks_progress/ Poll the progress of metric optimization from feedback processing. Returns the improved metric description and evaluation trigger when complete. # Update Metric Review Source: https://docs.cekura.ai/api-reference/test_framework/metric-review-update patch /test_framework/metric-reviews/{id}/ Update a metric review in Labs. Typically used to add or update feedback notes for test sets. # Process Feedbacks Answer Source: https://docs.cekura.ai/api-reference/test_framework/metric-reviews-process-feedbacks-answer post /test_framework/metric-reviews/process_feedbacks_answer/ Answer the questions a running metric optimization is paused on. Answers are keyed by question id; use {"action": "skip"} to skip one. # Get Metric Review Progress Source: https://docs.cekura.ai/api-reference/test_framework/metric-run-reviews-progress get /test_framework/v1/metrics/run_reviews_progress/ Poll the progress of metric re-evaluation. Returns current state and full metric data when complete. > [!WARNING] > > Deprecated. Use [Run Metric Reviews (v3)](/api-reference/test_framework/run-metric-reviews-v3) instead: it returns the result in one call, with no polling. # Move Folder Source: https://docs.cekura.ai/api-reference/test_framework/move-folder post /test_framework/v1/scenarios/move_folder/ Move a scenario folder to a new parent # Poll Async Metric Description Generation Progress Source: https://docs.cekura.ai/api-reference/test_framework/poll-async-metric-description-generation-progress get /test_framework/v1/metrics/generate_clean_description_progress/ Poll the result of a `generate_clean_description_bg` job. `status` is one of: `queued`, `processing`, `completed`, `failed`. On `completed`, `clean_description` and `thinking` are present. > [!WARNING] > > Deprecated. Use [Generate Clean Metric Description (v3)](/api-reference/test_framework/generate-clean-metric-description-v3) instead: it returns the result in one call, with no polling. # Poll Agent Auto-fetch Progress Source: https://docs.cekura.ai/api-reference/test_framework/poll-import-progress get /test_framework/v2/aiagents/{id}/auto-fetch-progress/ Poll the staged status of an agent auto-fetch (provider import). Each stage reports pending / in_progress / completed / skipped / failed. > [!WARNING] > > Deprecated. Use [Auto-fetch Agent from Provider (v3)](/api-reference/test_framework/re-import-agent-v3) instead: it returns the result in one call, with no polling. # Run Metric Reviews Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworkmetrics-run_reviews post /test_framework/v1/metrics/{id}/run_reviews/ Re-evaluate all test sets for this metric with optional overrides. Returns a progress_id to poll for completion. > [!WARNING] > > Deprecated. Use [Run Metric Reviews (v3)](/api-reference/test_framework/run-metric-reviews-v3) instead: it returns the result in one call, with no polling. # Generate shareable link for result Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworkresults-create_shareable_link_token post /test_framework/v1/results/{id}/create_shareable_link_token/ Generates a secure, time-limited shareable link that allows users to view test results without authentication. This feature requires an active subscription with whitelabel_reports enabled and provides both standard and custom branded URLs based on organization settings. # Delete runs from result Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworkresults-delete_runs post /test_framework/v1/results/{id}/delete_runs/ Soft deletes specified runs from a result by marking them as deleted and triggers result re-evaluation. # End all calls in result Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworkresults-end_calls post /test_framework/v1/results/{id}/end_calls/ Terminates all active calls associated with the runs in this result. This will end all ongoing phone calls or chat sessions for all non-deleted runs. # End a single call Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworkruns-end_call post /test_framework/v1/runs/{id}/end_call/ Terminates the active call associated with this specific run. This will end the ongoing phone call or chat session for this run. # Create Test Set from Call Log Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworktest-setscreate_from_call_log post /test_framework/test-sets/create_from_call_log/ Create a test set from call log data # Create Test Set from Run Source: https://docs.cekura.ai/api-reference/test_framework/post-test_frameworktest-setscreate_from_run post /test_framework/test-sets/create_from_run/ Create a test set from test run data # Preview Metric V3 Source: https://docs.cekura.ai/api-reference/test_framework/preview-metric-v3 post /test_framework/v3/metrics/preview/ Evaluate a metric configuration on up to 10 call logs or runs and return the verdicts, without saving the metric. Use it to check a metric before creating or updating it. Usually 30 seconds to 3 minutes; custom-code metrics can take longer. # Reevaluate Metrics Source: https://docs.cekura.ai/api-reference/test_framework/reevaluate-metrics post /test_framework/v1/results/{id}/evaluate_metrics/ Evaluate metrics against result transcripts # Rename Folder Source: https://docs.cekura.ai/api-reference/test_framework/rename-folder post /test_framework/v1/scenarios/rename_folder/ Rename a scenario folder # Reorder External Metrics Source: https://docs.cekura.ai/api-reference/test_framework/reorder-external-metrics patch /test_framework/v1/metrics-external/reorder/ Reorder project metrics # Reorder Metrics Source: https://docs.cekura.ai/api-reference/test_framework/reorder-metrics patch /test_framework/v1/metrics/reorder/ Reorder project metrics # Reorder Metrics V2 Source: https://docs.cekura.ai/api-reference/test_framework/reorder-metrics-v2 patch /test_framework/v2/metrics/reorder/ Reorder project metrics # Rerun Results Source: https://docs.cekura.ai/api-reference/test_framework/rerun-results post /test_framework/v1/results/{id}/rerun/ Re-run a test result evaluation # Run Evaluator (API Chat) Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-api-chat post /test_framework/v1/scenarios/run_scenarios_api_chat/ Start API chat runs you drive turn-by-turn over REST. No chat provider needs to be configured. # Run Evaluator CHIRP Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-chirp post /test_framework/v1/scenarios/run_scenarios_chirp/ Execute evaluators against an agent that exposes a raw-PCM WebSocket endpoint (16 kHz, 16-bit mono). Requires a WebSocket URL on the agent (`telephony.websocket_url`), or `websocket_url` in the request. # Run Evaluator ElevenLabs Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-elevenlabs post /test_framework/v1/scenarios/run_scenarios_elevenlabs/ Execute evaluators against an ElevenLabs Conversational AI agent. Requires ElevenLabs credentials and `assistant_id` set to the ElevenLabs agent ID. # Run Evaluator Email Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-email post /test_framework/v1/scenarios/run_scenarios_email/ Execute evaluators against an email agent. Cekura sends an email to the agent and simulates the full conversation via inbound webhook. # Run Evaluator LiveKit (v2) Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-livekit-v2 post /test_framework/v1/scenarios/run_scenarios_livekit_v2/ Execute evaluators against a LiveKit agent using project-level credentials. Cekura creates the rooms — no per-scenario URLs needed. Requires LiveKit credentials configured on the agent or project. # Run Evaluator Pipecat Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-pipecat post /test_framework/v1/scenarios/run_scenarios_pipecat/ Execute evaluators against an agent running on Pipecat / Daily.co WebRTC. Each scenario entry carries its own `pipecat_room_url` and optional token. # Run Evaluator Pipecat Cloud (v2) Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-pipecat-v2 post /test_framework/v1/scenarios/run_scenarios_pipecat_v2/ Execute evaluators against an agent deployed on Pipecat Cloud using project-level credentials. Cekura creates the sessions — no per-scenario room URLs needed. Requires Pipecat Cloud credentials configured on the agent or project. # Run Evaluator Retell WebRTC Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-retell-webrtc post /test_framework/v1/scenarios/run_scenarios_retell_webrtc/ Execute evaluators against a Retell agent over WebRTC (LiveKit infrastructure). Requires Retell credentials configured on the agent. # Run Evaluator SIP Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-sip post /test_framework/v1/scenarios/run_scenarios_sip/ Execute evaluators against an agent over a direct SIP connection (orchestrated by Cekura's SIP voice service). Requires `sip_endpoint` configured on the agent. # Run Evaluator Text Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-text post /test_framework/v1/scenarios/run_scenarios_text/ Execute evaluators against your agent using text-only simulation (SMS, chat, or an agent-hosted websocket). Cheaper than voice — billed at the text-simulation rate. # Run Evaluator V2 Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-v2 post /test_framework/v2/scenarios/run/ Start evaluator runs against an agent. Pick how the simulated user reaches the agent with `transport`; it must be one of the agent's `enabled_connections` (see `aiagents_retrieve`), and a transport that is not enabled is rejected with the list of enabled ones. Fields that belong to another transport are rejected. # Run Evaluator VAPI WebRTC Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-vapi-webrtc post /test_framework/v1/scenarios/run_scenarios_vapi_webrtc/ Execute evaluators against a VAPI agent over WebRTC. Requires VAPI credentials configured on the agent; `vapi_assistant_id` defaults to the agent's assistant ID. # Run Evaluator Voice Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-voice post /test_framework/v1/scenarios/run_scenarios/ Execute evaluators against your agent using a real voice / phone / SIP call. This is the primary 'run your evaluators' endpoint for voice-mode testing. # Run Evaluator Websocket Source: https://docs.cekura.ai/api-reference/test_framework/run-evaluator-websocket post /test_framework/v1/scenarios/run_scenarios_with_websockets/ Execute evaluators where each scenario carries its own websocket URL. Useful when you have a fleet of agents or stateful websocket sessions to test in a single batch. # Run Metric Reviews V3 Source: https://docs.cekura.ai/api-reference/test_framework/run-metric-reviews-v3 post /test_framework/v3/metrics/{id}/run_reviews/ Re-evaluate every review of the metric, optionally with overridden settings, and return each verdict next to its expected value. With `skip_save` the reviews are left unchanged. Each review is billed as a metric evaluation. Usually takes under a minute. # Run Scenarios From Spec JSON Source: https://docs.cekura.ai/api-reference/test_framework/run-scenarios-from-spec-json post /test_framework/v1/scenarios/run_scenarios_json/ Run agent test cases straight from a JSON spec, without creating evaluators. Keep the spec in your repository and post it from CI. # Poll Improvement Progress Source: https://docs.cekura.ai/api-reference/test_framework/scenario-agent-progress get /test_framework/v1/scenarios/scenario-agent-progress/ Check the progress of an evaluator improvement request. # Simplify Metric Prompt with AI Source: https://docs.cekura.ai/api-reference/test_framework/simplify-metric-prompt post /test_framework/v1/metrics/simplify_prompt/ Simplify a verbose metric prompt into a minimal, algorithmic version. Returns immediately if the metric is already optimized, otherwise returns a `progress_id` for polling via `simplify_prompt_progress/`. > [!WARNING] > > Deprecated. Use [Simplify Metric Prompt with AI (v3)](/api-reference/test_framework/simplify-metric-prompt-v3) instead: it returns the result in one call, with no polling. # Poll Simplify Metric Prompt Progress Source: https://docs.cekura.ai/api-reference/test_framework/simplify-metric-prompt-progress get /test_framework/v1/metrics/simplify_prompt_progress/ Poll the progress of an async metric prompt simplification job started by `simplify_prompt/`. > [!WARNING] > > Deprecated. Use [Simplify Metric Prompt with AI (v3)](/api-reference/test_framework/simplify-metric-prompt-v3) instead: it returns the result in one call, with no polling. # Simplify Metric Prompt V3 Source: https://docs.cekura.ai/api-reference/test_framework/simplify-metric-prompt-v3 post /test_framework/v3/metrics/simplify_prompt/ Rewrite a verbose metric prompt into a minimal version focused on its decision logic. When `metric_id` names an already-optimized metric whose prompt is unchanged, the stored prompt comes back at once. Usually takes under 30 seconds. # Start Async Metric Description Generation Source: https://docs.cekura.ai/api-reference/test_framework/start-async-metric-description-generation post /test_framework/v1/metrics/generate_clean_description_bg/ Enqueue a background job to generate a clean metric description. Returns a `progress_id`; poll `generate_clean_description_progress` until `status` is `completed` or `failed`. > [!WARNING] > > Deprecated. Use [Generate Clean Metric Description (v3)](/api-reference/test_framework/generate-clean-metric-description-v3) instead: it returns the result in one call, with no polling. # Update Agent (Full) Source: https://docs.cekura.ai/api-reference/test_framework/update-agent-full put /test_framework/v2/aiagents/{id}/ Fully replace an AI agent's fields. Prefer `PATCH` (partial_update) unless you need full-replacement semantics. See `aiagents_create` for the field shape and per-provider examples. # Update Agent (Partial) Source: https://docs.cekura.ai/api-reference/test_framework/update-agent-partial patch /test_framework/v2/aiagents/{id}/ Partially update an AI agent. Same field shape as create — see `aiagents_create` for details and per-provider examples. # Update Critical Metric Scenario Source: https://docs.cekura.ai/api-reference/test_framework/update-critical-metric-scenario patch /test_framework/v1/critical-metric-scenarios/{id}/ Update a critical metric scenario flag # Update Evaluator Source: https://docs.cekura.ai/api-reference/test_framework/update-evaluator patch /test_framework/v1/scenarios/{id}/ Partially update an evaluator. Same field shape as create — see `scenarios_create` for details. # Update External Scenario Condition Audio Source: https://docs.cekura.ai/api-reference/test_framework/update-external-scenario-condition-audio put /test_framework/v1/scenarios-external/{id}/condition-audio/{audio_id}/ Replace this clip's audio. Every action referencing it plays the new recording. Pass `name` to rename the clip at the same time — its `