Honeyhive
for LuumenAI
Read HoneyHive runs, events, and datasets, then change them as approved steps
Connect HoneyHive and your agent can read the evaluation history behind an AI application: runs and their metrics, event trees for a single session, datasets, and configurations. When something needs to change — a failing case added to a regression dataset, a new configuration, a batch of logged events — the agent shows the plan first and waits for your approval. Authentication is by API key.
The Honeyhive toolbox
42 tools: 19 read, 23 write. Reads answer instantly. Writes require approval by default. Everything is logged.
- ReadCompare Experiment RunsRetrieve experiment comparison between two evaluation runs.
- ReadCompare Runs EventsCompare events between two experiment runs side-by-side.
- ReadGet ConfigurationsRetrieve a list of configurations.
- ReadGet DatasetsRetrieve datasets from HoneyHive for a specified project.
- ReadGet EventsQuery events with filters and projections from HoneyHive.
- ReadGet Events By Session IDRetrieve the complete tree of nested events for a specific session.
- ReadGet Events ChartRetrieve charting and analytics data for events over time.
- ReadGet MetricsRetrieves all metrics associated with a HoneyHive project.
- ReadGet ProjectsRetrieve all projects in the HoneyHive account.
- ReadGet Evaluation Run DetailsGet details of an evaluation run by its UUID.
- ReadGet Run MetricsGet event metrics for an experiment run.
- ReadGet Evaluation RunsRetrieve a list of evaluation runs from HoneyHive.
- ReadGet Runs SchemaRetrieve the schema for experiment runs in HoneyHive.
- ReadGet SessionRetrieve a complete session tree by session ID from HoneyHive.
- ReadList ToolsList all available Honeyhive tools.
- ReadRetrieve DatapointRetrieve a specific datapoint by its ID from HoneyHive.
- ReadRetrieve DatapointsRetrieve datapoints from a HoneyHive project.
- ReadRetrieve EventsRetrieve and export events from a HoneyHive project.
- ReadRetrieve Experiment ResultRetrieve the result of a specific experiment run.
- WriteAdd datapoints to datasetAdd datapoints to a dataset. Approval by default
- WriteBatch Create DatapointsCreate multiple datapoints in a single batch operation. Approval by default
- WriteCreate Batch Model EventsCreate multiple model events in a single request. Approval by default
- WriteCreate Batch Tool EventsLog a batch of external API calls as tool events. Approval by default
- WriteCreate ConfigurationCreates a new configuration in HoneyHive for managing LLM or pipeline settings. Approval by default
- WriteCreate DatapointCreate a new datapoint with input-output pairs. Approval by default
- WriteCreate DatasetCreate a dataset. Approval by default
- WriteCreate EventCreate a new event in HoneyHive to track execution of different parts of your application. Approval by default
- WriteCreate MetricCreate a new metric in HoneyHive. Approval by default
- WriteCreate Model EventCreate a new model event to log LLM call data. Approval by default
- WriteCreate ToolCreates a new tool definition in a HoneyHive project. Approval by default
- WriteDelete DatapointDelete a specific datapoint by its ID. Approval by default
- WriteDelete DatasetDelete a dataset by ID. Approval by default
- WriteEnd Evaluation RunUpdate an evaluation run's status and metadata. Approval by default
- WriteStart Evaluation RunCreates a new evaluation run to group and track multiple session events for analysis. Approval by default
- WriteStart SessionStart a new HoneyHive session for tracing and observability. Approval by default
- WriteUpdate ConfigurationUpdate an existing HoneyHive configuration. Approval by default
- WriteUpdate DatapointUpdate an existing datapoint by ID. Approval by default
- WriteUpdate DatasetUpdate an existing dataset. Approval by default
- WriteUpdate EventUpdate an existing HoneyHive event by ID. Approval by default
- WriteUpdate MetricUpdate an existing metric. Approval by default
- WriteUpdate ProjectUpdates an existing HoneyHive project's name or description. Approval by default
- WriteUpdate ToolUpdate an existing tool in HoneyHive. Approval by default
One prompt, start to finish
What a governed Honeyhive run looks like inside Luumen.
Questions
Authorize once with API token. Luumen lists the scopes each action needs before you approve the connection, and credentials never appear in the chat.
Read actions answer immediately. Anything that writes — add datapoints to dataset, batch create datapoints, create batch model events, create batch tool events, and more — is shown as a plan and requires approval by default, including the 2 actions classified as destructive. Administrators configure that per tool, so you decide exactly which actions can ever run unattended.
You decide. Actions are granted per agent, skill, and team, and per environment — production is not staging. Read access can be broad while writes stay narrow.
Every call to Honeyhive — read or write, approved or declined — is recorded with the actor, the input, and the result, and can be linked to the ticket or change record.
Often connected alongside
All integrations
Adobe Marketing Agent MCPAdobe Experience Platform context, sandbox switches, and async task status
AffindaRead parsed documents, collections, and annotations — change them with approval
Agent MailGive the agent its own mailbox for reading threads and sending updates
AI/ML APIInspect assistants, threads, and runs — and change them as approved stepsPut Honeyhive to work with Luumen
Connect in minutes. Every action scoped, approved, and audited from day one.