MCP ツールリファレンス

ツールはガバナンス切片と機能カテゴリでグループ化しています。各ツールには呼び出しに必要な scope を併記。FactVerse のソース定義から生成。実際に利用できるツールは MCP の実行時ディスカバリを基準にしてください。

概要

スライス	エンドポイント	必要な scope	ツール数
ベース — 共通基盤	`/mcp/base/`	`base.action.write`, `base.compute.run`, `base.read`	54
TrafficOps — 交通・検問	`/mcp/trafficops/`	`trafficops.read`	7
PdM — 予知保全	`/mcp/pdm/`	`pdm.read`	5
TelcoOps — 通信ネットワーク	`/mcp/telcoops/`	`telcoops.read`	3
SemiOps — 半導体・クリーンルーム	`/mcp/semiops/`	`semiops.read`	16
Aviation — 航空信頼性	`/mcp/aviation/`	`aviation.analysis.read`	6

ベース — 共通基盤 · /mcp/base/ · 54 ツール数

エンドポイント /mcp/base/ · 必要な scope: base.action.write, base.compute.run, base.read

実行（書き込み）

create_work_order · base.action.write — Create a new work order based on the advisor's recommendation. Only use this when the user explicitly agrees to take action.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_id`	integer	はい	—	Equipment to create work order for
`title`	string	はい	—	Work order title
`description`	string	はい	—	Detailed description of work needed
`priority`	string	はい	—	(`LOW`, `MEDIUM`, `HIGH`, `CRITICAL`)

get_pending_tasks · base.read — Get pending ECM workflow tasks (documents awaiting approval, signature, or review). Use when the user asks about their to-do list or pending approvals.

パラメータ

パラメータ	型	必須	デフォルト	説明
`user_id`	integer	—	—	User ID (omit for current user)

分析

analyze_spare_parts · base.compute.run — Analyze spare parts inventory and usage patterns. Shows top replaced parts, slow-moving inventory, stockout risks, and reorder recommendations. Use for inventory optimization and procurement planning.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_type`	string	—	—	Filter by equipment type (e.g. AHU, CHILLER)
`part_category`	string	—	—	Filter by part category
`months`	integer	—	`12`	Months of history to analyze

calculate_emissions · base.compute.run — Calculate GHG emissions (Scope 1/2/3) for a building or facility. Returns CO₂e breakdown by scope with regional default default emission factors. Use when the user asks about carbon footprint, emissions, or sustainability metrics.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scope`	integer	はい	—	Emission scope: 1=direct, 2=electricity, 3=value chain (`1`, `2`, `3`)
`fuel_type`	string	—	—	Fuel type for Scope 1 (e.g. natural_gas, diesel, refrigerant_r410a)
`electricity_kwh`	number	—	—	Electricity consumption in kWh for Scope 2
`category`	string	—	—	Category for Scope 3 (e.g. waste_landfill, water_supply, commuting_mrt)
`consumption`	number	—	—	Consumption quantity in the relevant unit
`period`	string	—	—	Reporting period (e.g. '2025-01', '2025-Q1', '2025')

conformal_predict · base.compute.run — Distribution-free prediction intervals

宣言されたパラメータなし（実行時に tools/list で取得）。

detect_drift · base.compute.run — Detect data or concept drift between datasets

宣言されたパラメータなし（実行時に tools/list で取得）。

estimate_causal_effect · base.compute.run — Estimate treatment effects via causal inference

宣言されたパラメータなし（実行時に tools/list で取得）。

explain_prediction · base.compute.run — Explain a model prediction using SHAP

宣言されたパラメータなし（実行時に tools/list で取得）。

find_changepoints · base.compute.run — Detect structural breaks in a time series

宣言されたパラメータなし（実行時に tools/list で取得）。

fit_distribution · base.compute.run — Fit probability distributions to observed data. Useful for analyzing failure times, service durations, or arrival patterns. Returns best-fit distribution with parameters and goodness-of-fit statistics (KS test, AIC). Use when the user wants to understand what statistical distribution best describes their data.

パラメータ

パラメータ	型	必須	デフォルト	説明
`data_source`	string	はい	—	Source of data to fit. 'custom' expects raw data array. (`sensor_readings`, `failure_times`, `service_times`, `custom`)
`equipment_id`	integer	—	—	Equipment ID for sensor/failure data (required for sensor_readings, failure_times)
`sensor_type`	string	—	—	Sensor type filter (e.g. 'temperature', 'vibration') for sensor_readings
`custom_data`	array	—	—	Raw data points for custom fitting (min 20 points)

import_data · base.compute.run — Import and process data from external sources (REST API, CSV) through ETL pipeline for analysis. Connects to a data source, extracts records, and optionally fits distributions to the imported data. Use when the user wants to bring in external data for simulation input modeling.

パラメータ

パラメータ	型	必須	デフォルト	説明
`connector_type`	string	はい	—	Type of data connector: 'rest' for REST API, 'csv' for CSV file (`rest`, `csv`)
`endpoint`	string	はい	—	URL for REST API or file path for CSV
`pipeline_id`	string	—	—	Optional ETL pipeline: 'arrival-fitting' or 'service-time' (`arrival-fitting`, `service-time`)
`field_mapping`	object	—	—	Optional source→target field mapping (e.g. {'timestamp': 'arrival_time'})

データ・コネクタ

predict_surrogate · base.compute.run — Run inference with a trained surrogate model

宣言されたパラメータなし（実行時に tools/list で取得）。

recommend_model · base.compute.run — AutoML recommendation for best model type

宣言されたパラメータなし（実行時に tools/list で取得）。

train_surrogate · base.compute.run — Train a fast surrogate model from data

宣言されたパラメータなし（実行時に tools/list で取得）。

予測

automl_forecast · base.compute.run — AutoML model selection and forecast — picks the best algorithm

宣言されたパラメータなし（実行時に tools/list で取得）。

detect_anomaly · base.compute.run — Score data points for anomalies (z-score, isolation forest, autoencoder)

宣言されたパラメータなし（実行時に tools/list で取得）。

forecast_timeseries · base.compute.run — Run Holt-Winters or Prophet forecast on a time series

宣言されたパラメータなし（実行時に tools/list で取得）。

predict_rul · base.compute.run — Predict remaining useful life (RUL) from sensor readings

宣言されたパラメータなし（実行時に tools/list で取得）。

ナレッジ・参照

check_data_quality · base.read — Check data quality dashboard for all integrated data sources. Returns quality scores by dimension (completeness, accuracy, consistency, timeliness) and lists top violations. Use when asked about data quality, data health, or data issues.

宣言されたパラメータなし（実行時に tools/list で取得）。

extract_maintenance_record · base.read — Extract structured data from maintenance document (PDF/image) using AI. Returns equipment ID, maintenance date, type, technician, findings, replaced parts, and confidence score. Use when the user wants to digitize paper maintenance records.

パラメータ

パラメータ	型	必須	デフォルト	説明
`document_id`	integer	はい	—	ECM document ID to extract

get_action_plan_history · base.read — Get the history of AI action plans including their workflow approval status. Returns past action plans with approval decisions, execution outcomes, and linked ECM documents (incident reports). Use this to answer questions about past incidents, decisions, and their outcomes.

パラメータ

パラメータ	型	必須	デフォルト	説明
`checkpoint_id`	string	—	—	Optional: filter by checkpoint ID
`urgency`	string	—	`"ALL"`	Filter by urgency level. Default ALL (`CRITICAL`, `WARNING`, `ALL`)
`limit`	integer	—	`10`	Maximum number of results to return. Default 10

get_compliance_documents · base.read — Get compliance-related documents (certificates, audit reports, evidence packs) filtered by standard. Use when the user asks about ISO compliance, FDA, or regulatory documentation.

パラメータ

パラメータ	型	必須	デフォルト	説明
`standard`	string	—	—	Compliance standard (e.g. ISO_14644, SEMI_S2, GM, FDA_21_CFR)
`status`	string	—	—	Filter by document status (`APPROVED`, `EXPIRED`, `IN_REVIEW`, `RECORD`)

get_equipment_documents · base.read — Get all documents associated with an equipment (manuals, SOPs, drawings, maintenance records). Use when the user asks about equipment documentation or wants to find related manuals/SOPs.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_id`	integer	はい	—	Equipment ID
`doc_type`	string	—	—	Filter by document type (`MANUAL`, `SOP`, `DRAWING`, `REPORT`, `CERTIFICATE`, `PHOTO`)

get_equipment_status · base.read — Get real-time status of equipment including latest sensor readings, active alerts, and recent work orders. Use this to understand current conditions.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_id`	integer	—	—	Equipment ID to query (omit for all equipment)

get_expiring_documents · base.read — Get documents approaching retention end date or requiring periodic review. Use for compliance monitoring and proactive document management.

パラメータ

パラメータ	型	必須	デフォルト	説明
`days_ahead`	integer	—	`30`	Days ahead to check (default: 30)

list_connectors · base.read — List all configured data connectors (REST, CSV, MQTT, OPC-UA, database, etc.) with their current status (active/error/syncing), last sync time, and source info. Use when asked about data sources, integrations, connectors, or data pipelines.

宣言されたパラメータなし（実行時に tools/list で取得）。

query_knowledge · base.read — Query the knowledge graph for equipment types, failure modes, repair actions, diagnostic rules, and maintenance schedules. Use this to find expert knowledge.

パラメータ

パラメータ	型	必須	デフォルト	説明
`query_type`	string	はい	—	Type of knowledge to query (`equipment_info`, `failure_modes`, `repair_actions`, `diagnostic_rules`)
`equipment_type`	string	—	—	Equipment type (e.g. COMPRESSOR, AHU, PUMP, CHILLER)
`keyword`	string	—	—	Search keyword for free-text knowledge search

recommend_training · base.read — Recommend training courses based on equipment type, user role, and identified skill gaps. Returns prioritized course list with duration and priority level. Use for workforce development and certification planning.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_type`	string	—	—	Equipment type (e.g. AHU, CHILLER, COMPRESSOR)
`user_role`	string	—	`"operator"`	User role for role-specific training (`operator`, `technician`, `engineer`, `manager`)
`skill_gap`	string	—	—	Identified skill gap to address

search_checkpoint_sop · base.read — Search checkpoint Standard Operating Procedures (SOPs) stored in the ECM system. Returns relevant SOP documents for a given checkpoint or operation type. Use this when the user asks about procedures, protocols, or standard operations for checkpoint management, immigration control, or customs clearance. Leverages ECM RAG (Retrieval-Augmented Generation) for semantic search.

パラメータ

パラメータ	型	必須	デフォルト	説明
`query`	string	はい	—	Search query for SOP content (e.g. 'peak hour lane opening procedure', 'biometric scanner fallback protocol', 'VIP passenger handling')
`checkpoint_id`	string	—	—	Optional: specific checkpoint ID to scope the search
`doc_type`	string	—	`"ALL"`	Type of document to search. Default ALL (`SOP`, `INCIDENT_REPORT`, `CAPACITY_PLANNING`, `ALL`)

search_documents · base.read — Search ECM (Enterprise Content Management) for documents by keyword, type, or related entity. Returns document title, version, classification, and direct link. With RAG enhancement enabled, also provides AI-generated summary of relevant documents. Use when the user asks about manuals, SOPs, reports, certificates, or any documentation.

パラメータ

パラメータ	型	必須	デフォルト	説明
`query`	string	—	—	Search keyword (title, description, content)
`doc_type`	string	—	—	Filter by document type (`MANUAL`, `SOP`, `REPORT`, `DRAWING`, `CERTIFICATE`, `PHOTO`, `CONTRACT`, `TEMPLATE`)
`entity_type`	string	—	—	Filter by related entity type (`EQUIPMENT`, `ALERT`, `WORK_ORDER`, `CLEANROOM`, `SMT_LINE`)
`entity_id`	integer	—	—	Entity ID to filter by
`use_rag`	boolean	—	`true`	Enable RAG-based AI summary

troubleshoot_connector · base.read — Diagnose a specific data connector by fetching its details and recent sync logs. Analyzes recent errors and suggests concrete fixes (credentials, network, schema mapping, etc.). Use when a connector is failing, data is not syncing, or the user reports import/export issues.

パラメータ

パラメータ	型	必須	デフォルト	説明
`connector_name`	string	はい	—	Name or partial name of the connector to troubleshoot

最適化

find_optimal_policy · base.compute.run — Find optimal treatment policy using causal inference

宣言されたパラメータなし（実行時に tools/list で取得）。

get_optimization_recommendation · base.compute.run — Find the optimal staff configuration for checkpoint operations within a budget constraint. Uses NSGA-II multi-objective optimization to balance throughput vs wait time. Returns Pareto-optimal solutions with cost-benefit analysis and operator-ready actions. Use this to answer questions like 'How should we allocate 10 extra staff?' or 'What's the best configuration for a $5000/hr budget?'.

パラメータ

パラメータ	型	必須	デフォルト	説明
`budget`	number	—	`5000`	Total hourly budget for staff (cost units). Default 5000
`cost_per_staff`	number	—	`100`	Cost per additional staff member per hour. Default 100
`target_kpi`	string	—	`"avg_wait"`	Primary KPI to optimize (`avg_wait`, `throughput`, `p95_wait`)
`audience`	string	—	`"both"`	Target audience for the insight report (`manager`, `operator`, `both`)

optimize_bayesian · base.compute.run — Bayesian optimization for black-box function tuning

宣言されたパラメータなし（実行時に tools/list で取得）。

optimize_evolutionary · base.compute.run — Evolutionary multi-objective optimization (NSGA-II)

宣言されたパラメータなし（実行時に tools/list で取得）。

optimize_layout · base.compute.run — Optimize the spatial layout of a facility (checkpoint positions, capacities, routes) using NSGA-II multi-objective optimization with DES evaluation. Finds Pareto-optimal layouts balancing throughput vs wait time. Use for facility design and space planning.

パラメータ

パラメータ	型	必須	デフォルト	説明
`template_id`	string	はい	—	Layout template to optimize (`immigration-hall-small`, `security-screening`, `departure-lounge`)
`objectives`	array	—	—	Objectives to optimize (default: throughput, avg_wait_time)
`pop_size`	integer	—	—	NSGA-II population size (default: 20)
`n_gen`	integer	—	—	Number of generations (default: 30)

optimize_milp · base.compute.run — Solve a mixed-integer linear programming (MILP) problem

宣言されたパラメータなし（実行時に tools/list で取得）。

run_optimization · base.compute.run — Find optimal parameters using multi-objective optimization (NSGA-II). Returns Pareto-optimal solutions trading off competing objectives. Use this when the user wants to find the best configuration.

パラメータ

パラメータ	型	必須	デフォルト	説明
`module_type`	string	はい	—	(`trafficops`, `heatops`, `fms`)
`population_size`	integer	—	`20`
`generations`	integer	—	`10`

レポート

generate_report · base.compute.run — Generate a status report or simulation report. Use for equipment overview, alert summary, or simulation analysis reports.

パラメータ

パラメータ	型	必須	デフォルト	説明
`report_type`	string	はい	—	Type of report: 'simulation' runs a DES and reports KPIs, 'equipment_status' summarizes current equipment/alerts/work orders (`simulation`, `equipment_status`)
`module`	string	—	—	Module for simulation reports (`trafficops`, `heatops`, `fms`)
`format`	string	—	—	Output format (default: pdf) (`pdf`, `excel`)

シミュレーション

cascade_simulation · base.compute.run — Multi-engine cascade simulation — chain DES, ABM, Monte Carlo

宣言されたパラメータなし（実行時に tools/list で取得）。

run_abm · base.compute.run — Run an agent-based crowd simulation

宣言されたパラメータなし（実行時に tools/list で取得）。

run_dag_simulation · base.compute.run — Run a DAG-routed DES simulation with advanced routing (shortest-queue, probability, condition). Returns throughput, bottleneck analysis, Sankey flow data, and AI recommendations. Use for complex multi-path checkpoint scenarios.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_id`	string	はい	—	DAG scene to simulate (`cp-immigration-dag`, `cp-security-dag`, `cp-multi-terminal`)
`simulation_time`	number	—	—	Simulation duration in minutes (default: 120)
`staff_count`	integer	—	—	Number of staff/lanes (affects capacity)

run_des · base.compute.run — Run a discrete event simulation for process/queue modelling

宣言されたパラメータなし（実行時に tools/list で取得）。

run_doe · base.compute.run — Run a Design of Experiments to identify which factors most significantly affect a target metric. Returns ANOVA analysis showing factor significance.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_type`	string	はい	—	(`trafficops`, `heatops`, `fms`)
`scene_id`	string	はい	—
`factors`	array	はい	—	Factors to vary in the experiment
`response_metric`	string	はい	—	KPI to analyze (e.g. throughput, availability, total_heat_delivered_kj)

run_montecarlo · base.compute.run — Monte Carlo stress test / risk simulation

宣言されたパラメータなし（実行時に tools/list で取得）。

run_simulation · base.compute.run — Run a discrete event simulation (DES) to test a what-if scenario. Use this to verify predictions, compare configurations, or estimate impact of changes. Available scenes: trafficops (checkpoint flow), heatops (district heating), fms (equipment lifecycle).

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_type`	string	はい	—	Type of simulation to run (`trafficops`, `heatops`, `fms`)
`scene_id`	string	はい	—	Scene configuration ID (e.g. 'rts-main-hall', 'small-network', 'hvac-fleet')
`simulation_time`	number	—	`480`	Simulation time in minutes
`config_overrides`	object	—	—	Override scene parameters (e.g. num_counters, supply_temp)

run_system_dynamics · base.compute.run — Run a system dynamics (stock-and-flow) simulation

宣言されたパラメータなし（実行時に tools/list で取得）。

run_what_if_comparison · base.compute.run — Compare the current checkpoint configuration against a modified scenario using DES simulation. Use this to answer questions like 'What if we add 2 staff to biometric-scan?' or 'What happens if the bag scanner fails every 4 hours?'. Returns side-by-side KPI comparison, cost-benefit analysis, and concrete operator actions. Supports staff changes, lane adjustments, and equipment failure injection.

パラメータ

パラメータ	型	必須	デフォルト	説明
`changes`	object	—	—	Per-checkpoint overrides: {checkpoint_id: {staff_count, mean_service_time, counters}}. Example: {'biometric-scan': {'staff_count': 4}, 'bag-scan': {'staff_count': 3}}
`failure_injection`	object	—	—	Per-checkpoint failure configs: {checkpoint_id: {mtbf, mttr}}. Example: {'biometric-scan': {'mtbf': 240, 'mttr': 15}} — scanner fails every 4h, 15min repair
`label`	string	—	`"Modified Scenario"`	Human-readable label for the modified scenario
`audience`	string	—	`"both"`	Target audience for the insight report (`manager`, `operator`, `both`)
`replications`	integer	—	`5`	Number of simulation replications (higher = more accurate, slower)

simulate_logistics · base.compute.run — Run AGV/forklift logistics simulation on a facility layout

宣言されたパラメータなし（実行時に tools/list で取得）。

空間

analyze_spatial_anomaly · base.compute.run — Analyze spatial heatmap sensor data for anomalies. Detects sensors with values more than N standard deviations from the mean. Use when asked about hot/cold spots, unusual readings, or spatial outliers.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_id`	string	はい	—	Scene ID (heatops, iaq-building-env, energy-floor-consumption, space-occupancy)
`variable`	string	はい	—	Variable to analyze (supply_temp, co2, electricity)
`zone`	string	—	`"all"`	Zone filter: all, north, south, etc.
`threshold_sigma`	number	—	`2`	Sigma threshold for anomaly detection (default 2.0)

compare_zones · base.compute.run — Compare spatial statistics between two zones or floors. Provides mean, min, max, std dev comparison with interpretation. Use when asked to compare north vs south, floor 1 vs floor 2, or any two zones.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_id`	string	はい	—	Scene ID
`variable`	string	はい	—	Variable to compare
`zone_a`	string	はい	—	First zone (north, 1f, etc.)
`zone_b`	string	はい	—	Second zone (south, 2f, etc.)

import_dxf · base.compute.run — Import a DXF floor plan and recognize walls/doors/windows/fences

宣言されたパラメータなし（実行時に tools/list で取得）。

recommend_sensor_placement · base.compute.run — Recommend optimal locations for additional sensors based on spatial coverage gaps and IDW confidence analysis. Use when asked about sensor deployment, coverage gaps, or where to install new sensors.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_id`	string	はい	—	Scene ID
`variable`	string	はい	—	Variable to analyze for coverage
`zone`	string	—	`"all"`	Zone to analyze. Use all for full area.
`max_recommendations`	integer	—	`5`	Maximum number of placement recommendations (default 5)

経路案内

find_path · base.compute.run — Find navigation path between two locations in a building. Supports multi-floor routing, accessibility options, and crowd avoidance.

パラメータ

パラメータ	型	必須	デフォルト	説明
`from_location`	string	はい	—	Starting location name or node ID
`to_location`	string	はい	—	Destination location name or node ID
`accessible`	boolean	—	`false`	Wheelchair-accessible route only
`avoid_crowds`	boolean	—	`false`	Avoid congested areas

TrafficOps — 交通・検問 · /mcp/trafficops/ · 7 ツール数

エンドポイント /mcp/trafficops/ · 必要な scope: trafficops.read

check_officer_roster · trafficops.read — Check current shift officer roster and manpower availability at the checkpoint. Returns all deployed officers with their assignments, available spares, and next shift change time. Use this when asked about staffing, manpower, or whether additional officers can be deployed.

パラメータ

パラメータ	型	必須	デフォルト	説明
`shift`	string	—	`"current"`	Shift to check: 'current', 'morning', 'afternoon', 'night'

evaluate_lane_reconfig · trafficops.read — Run a DES (Discrete Event Simulation) comparing the current lane configuration against a proposed reconfiguration (e.g., closing a car lane to open an additional motorcycle lane). Uses real simulation engine to compute wait times, throughput, and SLA compliance. You can provide user-reported data like arrival rates and queue lengths for accurate simulation. Use this when recommending lane changes to quantify the trade-off before the officer decides.

パラメータ

パラメータ	型	必須	デフォルト	説明
`close_lanes`	array	はい	—	Lane IDs to close (e.g. ['CAR-4'])
`open_lanes`	array	はい	—	New lane configs to open (e.g. [{'id': 'MC-6', 'type': 'motorcycle', 'from_lane': 'CAR-4'}])
`motorcycle_arrival_rate_hr`	number	—	—	Motorcycle arrival rate in vehicles/hour (default: 420 for surge scenario)
`car_arrival_rate_hr`	number	—	—	Car arrival rate in vehicles/hour (default: 180)
`motorcycle_lanes`	integer	—	—	Current number of motorcycle lanes in baseline (default: 5)
`car_lanes`	integer	—	—	Current number of car lanes in baseline (default: 4)
`motorcycle_queue_length`	integer	—	—	Current motorcycle queue length (total vehicles waiting, default: 0)
`car_queue_length`	integer	—	—	Current car queue length (total vehicles waiting, default: 0)
`simulation_time_min`	number	—	—	Simulation duration in minutes (default: 60)

get_checkpoint_lane_status · trafficops.read — Get real-time lane status for a vehicle checkpoint including per-lane utilization, queue lengths, wait times, and assigned officers. Covers both motorcycle and car lanes. You can provide user-reported data (queue lengths, arrival rates, lane counts) to override defaults. Use this when asked about current checkpoint conditions, congestion, or lane capacity.

パラメータ

パラメータ	型	必須	デフォルト	説明
`scene_id`	string	—	`"border-lbc-arrival-car"`	Scene ID (e.g. border-lbc-arrival-car, border-lbc-departure-car)
`motorcycle_lanes`	integer	—	—	Override number of motorcycle lanes (default: 5)
`car_lanes`	integer	—	—	Override number of car lanes (default: 4)
`motorcycle_queue_total`	integer	—	—	User-reported total motorcycle queue length across all lanes
`car_queue_total`	integer	—	—	User-reported total car queue length across all lanes
`motorcycle_arrival_rate_hr`	number	—	—	User-reported motorcycle arrival rate (vehicles/hour)
`car_arrival_rate_hr`	number	—	—	User-reported car arrival rate (vehicles/hour)

get_proactive_alerts · trafficops.read — Get proactive congestion alerts based on forecast vs SLA comparison. Returns predicted SLA breaches with severity, evidence, and improvement suggestions (Budget/Speed/Balanced). Use this when asked about potential upcoming issues or congestion risks.

パラメータ

パラメータ	型	必須	デフォルト	説明
`sla_minutes`	number	—	`12`	SLA threshold in minutes for wait time

get_surge_detection · trafficops.read — Get current traffic surge/anomaly detection status for a checkpoint. Detects unusual spikes in arrival rates by vehicle type (motorcycle, car, bus). Returns surge magnitude, estimated duration, probable cause, and initial recommendation. You can provide user-reported arrival rates to override defaults. Use this when asked about current traffic anomalies or unexpected congestion.

パラメータ

パラメータ	型	必須	デフォルト	説明
`checkpoint`	string	—	`"vehicle-arrival"`	Checkpoint area (e.g. vehicle-arrival, pedestrian-arrival)
`motorcycle_arrival_rate_hr`	number	—	—	User-reported motorcycle arrival rate (vehicles/hour)
`car_arrival_rate_hr`	number	—	—	User-reported car arrival rate (vehicles/hour)
`motorcycle_lanes`	integer	—	—	Number of motorcycle lanes (for affected lanes list)

get_traffic_forecast · trafficops.read — Get an 8-hour traffic flow forecast for a specific checkpoint. Returns predicted throughput (pax/h) at 15-minute intervals with confidence bands. Use this to check if congestion is expected and plan ahead.

パラメータ

パラメータ	型	必須	デフォルト	説明
`checkpoint`	string	—	`"document-check"`	Checkpoint ID (e.g. document-check, biometric-scan, security-screen)
`hours`	integer	—	`8`	Hours ahead to forecast (1-24)

get_traffic_patterns · trafficops.read — Get detected recurring patterns for a traffic checkpoint using DOE statistical analysis. Returns day-of-week effects, hour-of-day peaks, and bottleneck patterns with p-values and confidence levels. Use this to understand structural traffic behavior.

パラメータ

パラメータ	型	必須	デフォルト	説明
`checkpoint`	string	—	`"document-check"`	Checkpoint ID
`days`	integer	—	`90`	Days of historical data to analyze

PdM — 予知保全 · /mcp/pdm/ · 5 ツール数

エンドポイント /mcp/pdm/ · 必要な scope: pdm.read

get_equipment_health · pdm.read — Get health status of a predictive maintenance equipment. Returns health score (0-100), grade (A/B/C/D/F), anomaly level, crest factor, vibration RMS.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_name`	string	はい	—	Equipment name or code (e.g. VB-VP-001, vacuum pump)
`equipment_id`	string	—	—	Equipment UUID (optional)

get_filter_circular_recovery · pdm.read — Get live circular recovery and aftermarket outlook for predictive maintenance filter components in the current tenant or one equipment. Returns recovery candidates, risk band, remaining life, recommended actions, and aftermarket or disposal guidance. Use for circular recovery, reuse, remanufacture, disposal, aftermarket, or sustainability questions about filter components.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_id`	string	—	—	Equipment ID (optional)
`equipment_code`	string	—	—	Equipment code or name, e.g. MH-MP-001 (optional)
`top_n`	integer	—	—	How many components to highlight (default: 5)

get_filter_component_intelligence · pdm.read — Get live predictive maintenance filter component intelligence for the current tenant or one equipment. Returns components that need attention now, including risk band, predicted remaining life, recommended action, benchmark context, and aftermarket narrative. Use for questions about filter components, component intelligence, customer fleet filters, what needs attention right now, or component-level maintenance priorities.

パラメータ

パラメータ	型	必須	デフォルト	説明
`equipment_id`	string	—	—	Equipment ID (optional)
`equipment_code`	string	—	—	Equipment code or name, e.g. MH-EX-003 (optional)
`top_n`	integer	—	—	How many components to highlight (default: 5)

get_pdm_summary · pdm.read — Get predictive maintenance fleet health summary: equipment count, grade distribution (A/B/C/D/F), critical count, active alerts, 7-day health trend.

パラメータ

パラメータ	型	必須	デフォルト	説明
`tenant_filter`	string	—	—

list_pdm_anomalies · pdm.read — List predictive maintenance anomalies: bearing wear, overheating, vibration excess. Returns type, severity, equipment, AI recommendation.

パラメータ

パラメータ	型	必須	デフォルト	説明
`severity`	string	—	—	(`ALL`, `CRITICAL`, `HIGH`, `MEDIUM`, `LOW`)
`status`	string	—	—	(`OPEN`, `RESOLVED`, `ALL`)
`limit`	integer	—	—

TelcoOps — 通信ネットワーク · /mcp/telcoops/ · 3 ツール数

エンドポイント /mcp/telcoops/ · 必要な scope: telcoops.read

analyze_network_health · telcoops.read — Analyze telecom network health by fetching overview, link utilization, and open incidents. Returns a graded narrative (A-F) with node/link counts, incident summary, top-risk link, financial impact estimate, and prioritized recommendations. Use when asked about network status, NOC overview, telecom health, or infrastructure risk.

宣言されたパラメータなし（実行時に tools/list で取得）。

explain_incident · telcoops.read — Explain a specific telecom network incident in natural language. Returns severity, detection time, root cause, customer/revenue impact, and step-by-step remediation actions (immediate, short-term, long-term). Use when asked about a particular incident, alert, or fault.

パラメータ

パラメータ	型	必須	デフォルト	説明
`incident_id`	string	はい	—	Incident ID to explain

predict_capacity · telcoops.read — Predict link capacity breaches based on current utilization and growth trends. Identifies links above 70% utilization, estimates days to breach threshold, and returns CapEx requirements and SLA penalty exposure if upgrades are deferred. Use when asked about capacity planning, bandwidth forecasting, or upgrade priorities.

宣言されたパラメータなし（実行時に tools/list で取得）。

SemiOps — 半導体・クリーンルーム · /mcp/semiops/ · 16 ツール数

エンドポイント /mcp/semiops/ · 必要な scope: semiops.read

ナレッジ・参照

analyze_env_correlation · semiops.read — Analyze correlation between environmental parameters (temperature, humidity, pressure, particles) in a cleanroom. Helps identify which parameters influence each other. Especially useful for diagnosing temperature/humidity impact on solder paste performance, PCB lamination quality, and photolithography exposure accuracy.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	はい	—	Cleanroom ID to analyze

get_cleanroom_status · semiops.read — Get real-time cleanroom status including temperature, humidity, pressure, particle counts, and ISO compliance. Covers lamination rooms, PCB/FPC exposure zones, and general semiconductor cleanrooms. Use when asked about cleanroom conditions, environment, or contamination levels. Omit cleanroom_id to get all cleanrooms.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	—	—	Cleanroom ID (omit for all cleanrooms)

get_fab_pue · semiops.read — Get current Power Usage Effectiveness (PUE) for the fab facility with energy breakdown (IT load, cooling, lighting, HVAC, etc.) and benchmark rating. Applicable to PCB/FPC factories, semiconductor fabs, and electronics manufacturing plants. Use for energy efficiency questions.

宣言されたパラメータなし（実行時に tools/list で取得）。

get_filter_life · semiops.read — Get HEPA/ULPA filter remaining life prediction based on pressure drop trends. Shows estimated days remaining before filter replacement is needed. Covers cleanroom filters for PCB exposure areas, lamination zones, and semiconductor fabs. Do NOT use for predictive maintenance/mobile-equipment filter components such as excavators, loaders, generators, or customer fleet fleet assets.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	—	—	Cleanroom ID (omit for all filters)

get_iso_compliance · semiops.read — Get ISO 14644 compliance status and assessment history for cleanrooms. Shows current classification, pass/fail status, and historical assessment trends.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	—	—	Cleanroom ID (omit for all)

get_particle_trend · semiops.read — Get particle count trends over time for a specific cleanroom. Shows how particle levels have changed and helps identify contamination events or degradation patterns.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	はい	—	Cleanroom ID to query
`hours`	integer	—	`24`	Hours of history to retrieve
`particle_size`	string	—	—	Particle size filter (e.g. '0.5um', '5.0um')

get_pressure_gradient · semiops.read — Get pressure gradient cascade status across cleanroom pairs. Shows whether pressure differentials between rooms are maintained correctly to prevent cross-contamination.

宣言されたパラメータなし（実行時に tools/list で取得）。

get_smt_oee · semiops.read — Get SMT (Surface Mount Technology) production line OEE (Overall Equipment Effectiveness) with Availability × Performance × Quality breakdown. Covers PCB assembly lines including solder paste printing, pick-and-place, reflow oven, and AOI stations. Use for production efficiency questions.

パラメータ

パラメータ	型	必須	デフォルト	説明
`line_id`	string	—	—	SMT line ID (omit for all lines)

get_utility_status · semiops.read — Get utility systems status including CDA (Clean Dry Air), N2 (Nitrogen), PCW (Process Cooling Water), and UPW (Ultra Pure Water). Shows pressure, flow, purity readings.

宣言されたパラメータなし（実行時に tools/list で取得）。

モニタリング

monitor_particles · semiops.read — Monitor real-time particle counts in a cleanroom against ISO 14644-1 limits. Returns per-size evaluation, alerts for threshold exceedances, and overall status.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	はい	—	Cleanroom ID to monitor

run_soft_sensors · semiops.read — Run virtual soft sensors to estimate unmeasurable parameters (AMC molecular contamination in ppb, dew point °C, HEPA filter loading %) from available cleanroom sensor data. Use when asked about molecular contamination, AMC levels, dew point, or filter status and no direct measurement is available.

パラメータ

パラメータ	型	必須	デフォルト	説明
`temperature_c`	number	—	—	Current cleanroom temperature in °C
`humidity_pct`	number	—	—	Current relative humidity %
`particle_05um`	number	—	—	Current 0.5 µm particle count, particles/m³
`air_changes_hour`	integer	—	`600`	Air changes per hour (ACH)
`cleanroom_age_days`	integer	—	`365`	Age of the cleanroom in days (affects outgassing AMC estimate)
`filter_dp_pa`	number	—	`200`	Current HEPA/ULPA filter pressure drop in Pa
`filter_initial_dp_pa`	number	—	`50`	Pressure drop of a new clean filter in Pa
`filter_max_dp_pa`	number	—	`450`	Replacement-threshold filter pressure drop in Pa
`filter_operating_hours`	number	—	`4380`	Cumulative filter operating hours

最適化

optimize_chiller_cop · semiops.read — Optimize chiller loading across multiple units to maximize system COP. Compares optimal vs equal-loading strategies and calculates energy savings.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cooling_demand_kw`	number	はい	—	Total cooling demand in kW
`ambient_temp_c`	number	—	`35`	Outdoor ambient temperature °C

推論予測

forecast_fab_load · semiops.read — Forecast fab electrical load for the next 24-168 hours using pattern-based model. Identifies peak/valley periods and demand response opportunities.

パラメータ

パラメータ	型	必須	デフォルト	説明
`hours_ahead`	integer	—	`24`	Hours to forecast (1-168)

品質

classify_smt_defects · semiops.read — Classify SMT defects with Pareto analysis and root-cause recommendations. Shows defect distribution by type/severity, DPMO, and actionable fixes. Recognizes PCB/FPC-specific defect types including solder paste issues, tombstoning, bridging, missing components, cold joints, and pad lifting.

パラメータ

パラメータ	型	必須	デフォルト	説明
`line_id`	string	—	—	SMT line ID (optional, all lines if omitted)
`days`	integer	—	`7`	Days of defect data to analyze

シミュレーション

predict_env_trend · semiops.read — Predict environmental parameter trends (temperature, humidity, particles) for the next 2-4 hours in a cleanroom. Use for proactive monitoring and early warning.

パラメータ

パラメータ	型	必須	デフォルト	説明
`cleanroom_id`	string	はい	—	Cleanroom ID to predict
`parameter`	string	—	—	Environmental parameter to predict (`temperature`, `humidity`, `particles`, `pressure`)
`hours`	integer	—	`4`	Hours ahead to predict (1-8)

simulate_smt_bottleneck · semiops.read — Run discrete-event simulation of an SMT production line to identify throughput bottleneck station, utilization imbalances, and optimization opportunities.

パラメータ

パラメータ	型	必須	デフォルト	説明
`line_id`	string	—	—	SMT line ID (optional, uses default config)
`sim_hours`	number	—	`8`	Simulation duration in hours
`boards`	integer	—	`500`	Number of boards to produce

Aviation — 航空信頼性 · /mcp/aviation/ · 6 ツール数

エンドポイント /mcp/aviation/ · 必要な scope: aviation.analysis.read

aviation_component_compare · aviation.analysis.read — Compare reliability behavior across component groups.

宣言されたパラメータなし（実行時に tools/list で取得）。

aviation_fleet_stats · aviation.analysis.read — Summarize fleet reliability signals by ATA chapter and failure distribution.

宣言されたパラメータなし（実行時に tools/list で取得）。

aviation_kpi_attribution · aviation.analysis.read — Attribute a reliability KPI to supporting failure records and evidence.

宣言されたパラメータなし（実行時に tools/list で取得）。

aviation_repetitive_fault_detect · aviation.analysis.read — Detect and summarize repetitive fault groups for reliability review.

宣言されたパラメータなし（実行時に tools/list で取得）。

aviation_text_mining_scan · aviation.analysis.read — Scan maintenance text for recurring or unusual technical issue candidates.

宣言されたパラメータなし（実行時に tools/list で取得）。

aviation_weibull_fit · aviation.analysis.read — Fit Weibull reliability curves for selected replacement or removal data.

宣言されたパラメータなし（実行時に tools/list で取得）。

JSON リファレンス: tools.json