🧰Tool Reference

The full set of tools the agent can call against your BioBox graph — what each does, its parameters, and an example response.

When you ask a question, the agent chooses from the tools below and calls the ones it needs. You don't call these directly — the agent does — but this reference shows exactly what each tool does, what it accepts, and what it returns, so you can understand (and trust) how an answer was built.

These pages are generated automatically from the BioBox MCP tool definitions, so they always match what the agent actually sees. Every tool is read-only and scoped to your organization.

The tools, by what they're for

Get oriented in the graph

• list_entity_types — what kinds of things (genes, diseases, drugs, …) exist in the graph

• describe_entity_type — the properties and relationships of one entity type

• get_graph_schema — the overall shape of the graph

• get_concept_relationships — which ontology relationships can connect two entity types

• find_intermediate_concept — a bridging concept between two groups that don't connect directly

Find and inspect entities

• search_entities — resolve a name or symbol (e.g. "ERBB4") to graph nodes

• search_entities_fulltext — full-text search over entity names in the ontology

• search_entities_fuzzy — typo-tolerant search when the spelling isn't exact

• get_entity — full detail for one entity

• get_entity_connections — what an entity is connected to

• find_paths — the shortest path between two entities

• expand_anchor — broaden a too-specific entity up its is-a hierarchy to a parent with more data

• run_graph_query — a read-only graph query for advanced questions

Prioritization reports & evidence

• list_reports — the scoring reports available to you

• list_profile_reports — the profile (monopartite) scoring reports available to you

• get_report_scores — the ranked scores in a report

• get_evidence — the evidence behind a scored entity

• where_entity_ranks — where a specific entity ranks in a report

Compare reports (delta reports)

• list_delta_reports — the saved report-to-report comparisons available to you

• get_delta_report — the per-target diff (biggest movers) within a comparison

• get_delta_report_target — how one target changed between the compared runs

Reasoning models

• list_reasoning_models — the reasoning models in your organization, across both topologies

• get_reasoning_model — the configuration of one reasoning model, whichever topology it is

• search_models — find a reasoning model by name

Dashboards

• list_dashboards — the dashboards in your organization

• list_object_dashboards — the dashboards pinned to a specific entity

• get_dashboard — the widgets on one dashboard

• get_widget_data — run one dashboard widget's saved query and return its data

Query building & guidance

• run_bxql — run a read-only BxQL pattern query directly

• validate_bxql — check a BxQL query parses before relying on it

• ground_query — resolve a fuzzy mention to a real, data-backed value for a query

• get_ql_instructions — your organization's query-building instructions

• find_guidance — find guidance playbooks for a question's intent

• get_concept_guidance — the guidance playbooks attached to a concept

• list_query_templates — the reusable query templates your organization has

• get_query_template — one reusable query template by id

• list_evidence_catalog — your organization's catalog of canonical evidence lines

Browse any page for the details.