HoneyHive
Developer toolsAPI keyCredentials form
Search and log AI agent traces, manage evaluation datasets, and read experiment runs and configurations in HoneyHive.
13 actions
| Action | Effect | Data handled | Scope |
|---|---|---|---|
| honeyhive_search_eventsSearches trace events (sessions, model calls, tools, chains), newest first, by filters, a time range or an experiment run | read | Personal data | |
| honeyhive_get_eventGets one trace event with its inputs, outputs, config, metrics, feedback and metadata | read | Personal data | |
| honeyhive_create_eventLogs one trace event (model, tool, chain or session) with its inputs, outputs and metadata, for runtimes without the SDK | write | Personal data | |
| honeyhive_list_datasetsLists the project's datasets, or finds one by id or name | read | General | |
| honeyhive_create_datasetCreates a dataset, optionally holding existing datapoints | write | General | |
| honeyhive_list_datapointsLists datapoints, those of a dataset by name or those with given ids, with inputs and ground truth | read | Personal data | |
| honeyhive_create_datapointCreates a datapoint with inputs, ground truth and metadata, optionally linked to datasets or the event it came from | write | Personal data | |
| honeyhive_list_runsLists experiment runs, newest first by default, optionally by dataset, name or status | read | General | |
| honeyhive_get_runGets one experiment run with its configuration, dataset and status | read | General | |
| honeyhive_get_run_summarySummarizes an experiment run: each metric aggregated over its datapoints | read | General | |
| honeyhive_compare_runsCompares two experiment runs metric by metric | read | General | |
| honeyhive_list_configurationsLists prompt and model configurations, optionally by name, environment or tags, including which is deployed | read | General | |
| honeyhive_list_metricsLists the evaluators (metrics) defined in the project, optionally of one type | read | General |