Give agents a grounded sense of place.

A typed Model Context Protocol server and agent-ready APIs let AI systems discover, reason over, and act on Mappls safely—without putting secrets or invented endpoints into prompts.

Tools, not prompt folklore.

Every operation has a stable name, bounded schema, explicit auth scope, structured error, and provenance-bearing result.

01

Typed discovery

Agents can compose 208 scenario-platform plans and search 162 exact API contracts, 58 SDK/distribution records, 138 application event contracts, 13 platform quickstarts, 13 migrations, 39 tutorials, and the platform catalog at runtime.

02

Least privilege

Scopes and resource boundaries are enforced outside model context.

03

Human control

Sensitive writes can require review before execution.

04

Grounded output

Results retain Mappls IDs, inputs, region, source fingerprints, and lifecycle roles.

05

Operational safety

Timeouts, quotas, retries, and audit events are first-class.

06

Composable context

16 industry blueprints and 18 stateful journeys share the same product and contract vocabulary.

Small, precise, composable.

32 offline discovery, matrix, qualification, conformance, coverage, generation-comparison, solution, Academy, recipe, integration-manifest and agent planning, exact-source launchpad, journey simulation, environment-readiness, handoff, configuration, authentication, diagnostics, selection, compatibility, and detail tools ground implementation plans before ten read-only live tools touch Mappls services. Local stdio preserves the complete surface. Authenticated stateless HTTP defaults to the credential-free offline profile; a separately governed live-read deployment explicitly enables all 42. Dedicated capability-matrix, capability-qualification, live-conformance, integration-manifest, contract-generation, webhook-receiver, source-launchpad, documentation-coverage, implementation-recipe, learning-path, journey-simulation, agent-integration, environment, authentication, error, SDK, event-contract, stateful-planning, automotive, SkyDNN vision, repository-correspondence, and release-intelligence resources expose source boundaries, application-owned lifecycle rules, hostile-path proof, and honest unmatched evidence without protected artifacts, media, or credentials. No operation creates, dispatches, closes, or deletes an operational Mappls resource.

mappls_capability_catalog

Match a scenario to all ten Mappls product families, platforms, and state models.

offline
mappls_get_capability_matrix

Inspect any product-platform decision with exact evidence kept separate from adjacent learning and sample material.

offline
mappls_get_capability_qualification

Turn one product-platform decision into a six-phase, credential-free conformance and independent-release evidence journey.

offline
mappls_plan_live_conformance

Plan the eight allowlisted non-production reads, preserve all-ten-product boundaries, and emit the canonical credential-safe report contract without making a provider call.

offline
mappls_find_documentation

Search products, platforms, contracts, SDK references, migrations, tutorials, industries, journeys, and samples.

offline
mappls_plan_solution

Compose one deterministic industry-and-platform architecture with stateless lanes, durable journeys, evidence, and release gates.

offline
mappls_plan_learning_path

Personalize a beginner, intermediate, or advanced Developer Academy curriculum with code labs, hostile journeys, a tested capstone, and explicit exit evidence.

offline
mappls_get_implementation_recipe

Load one of 48 industry recipes at request-response, composed-experience, or durable-operation depth with eight-language code, recovery controls, canonical evidence, and a tested capstone.

offline
mappls_build_integration_manifest

Generate one of 624 portable use-case, platform, and operating-depth contracts with authentication generations, exact source fingerprints, journeys, runnable proof, and release gates.

offline
mappls_plan_stateful_integration

Generate the state-ownership boundary, aggregates, eight control layers, lifecycle evidence, runnable apps, and release gates for any journey-bearing product.

offline
mappls_simulate_journey

Execute all eighteen lifecycles through complete, replay, stale-version, invalid-transition, and unknown-outcome fixtures with immutable application event evidence.

offline
mappls_build_deep_link

Build a validated, credential-free place, navigation, point, or Pin hand-off from current official path evidence.

offline
mappls_build_widget_embed

Generate one of seven source-bounded widget iframes—including complete Add a Place and entitled RealView evidence journeys—or preserve the enterprise/native selection boundary.

offline
mappls_get_native_widget_path

Select an Android, iOS, Flutter, or React Native widget path with current versions, explicit gaps, lifecycle ownership, scaffold, and release tests.

offline
mappls_plan_mgis_integration

Separate the entitled mGIS widget, twelve Web JS methods, durable workspace APIs, and sovereign MIGIST with source and state boundaries.

offline
mappls_get_quickstart

Load one credential-safe first success with five proof-bearing stages, code, verification, and evidence boundaries.

offline
mappls_get_source_launchpad

Load one exact repository fingerprint, six-stage path, safe source examples, hostile recovery, release gates, and its offline integration harness.

offline
mappls_get_api_contract

Retrieve one normalized operation with auth, schemas, lifecycle role, source fingerprint, and samples.

offline
mappls_get_sdk

Retrieve one of 58 SDK or distribution decisions with install evidence, lifecycle, source identity, correspondence, migration paths, and compatibility gates.

offline
mappls_get_authentication_path

Choose or inspect one of 14 credential paths; ambiguous product/runtime pairs remain selection-required and no credential value is accepted.

offline
mappls_diagnose_error

Classify safe status/code evidence, choose bounded retry or reconcile-first, and produce a credential-safe support packet.

offline
mappls_plan_environment

Assess development, staging, or production against 11 conditional identity, access, traffic, event, owner, release, and rollback controls without accepting secrets or artifacts.

offline
mappls_plan_agent_integration

Compose one of 6 purpose-bound AI application patterns with an exact MCP allowlist, 12 conditional controls, 8 evaluation dimensions, durable-state ownership, and human control.

offline
mappls_get_migration

Retrieve a source-bounded cutover with candidate confidence, exit gates, code seam, acceptance, and rollback.

offline
mappls_get_tutorial

Load a complete progressive recipe with acceptance criteria and recovery guidance.

offline
mappls_get_industry_blueprint

Load a real-world architecture, durable data contracts, risks, KPIs, and rollout gates.

offline
mappls_get_journey_blueprint

Load states, commands, invariants, records, recovery, observability, and linked evidence.

offline
mappls_get_journey_workshop

Load the exact eight-lab implementation curriculum, five hostile scenarios, checksummed capstone, and production exit evidence for any durable journey.

offline
mappls_get_event_contract

Load one journey-scoped application event envelope, schema, producers, replay rules, consumer adapters, and its explicit provider-payload boundary.

offline
mappls_get_release_intelligence

Search all public sources by lifecycle, snapshot-relative activity, correspondence, fingerprint, release risk, and conservative compatibility action.

offline
mappls_get_documentation_coverage

Inspect exact, normalized, and adjacent documentation evidence plus prioritized gaps across every public source, product, and platform.

offline
mappls_get_contract_generation

Compare exact retained source generations, field-level changes, compatibility signals, fingerprints, and an application-owned migration checklist without inferring provider support.

offline
mappls_search_autosuggest

Resolve incomplete human place intent into verified Mappls places.

live read
mappls_search_nearby

Discover categorized places around a validated point or Mappls Pin.

live read
mappls_geocode

Normalize and locate a structured or free-form address.

live read
mappls_reverse_geocode

Explain a coordinate using Mappls address and place context.

live read
mappls_route

Plan a route with explicit profile, region, alternatives, and constraints.

live read
mappls_route_matrix

Compare travel cost across origins and destinations.

live read
mappls_asset_status

Read authorized live or last-known asset state without exposing raw credentials.

live read
mappls_asset_events

Read a bounded device-history window while preserving event-time evidence.

live read
mappls_geofence_activity

Inspect authorized entry and exit evidence without mutating an operational case.

live read
mappls_trip_status

Inspect an InTouch trip, route, device state, and optional events as one read model.

live read

From fixture to a real MCP resource.

The downloadable Spatial Operations Agent implements fixture, local stdio, and authenticated remote HTTP clients behind one governed plan/approve/execute contract. The new Agent Integration Center turns that implementation into six deployable authority patterns.

Offline solution research

Discover → architect → review → cite

16 industries × 13 platforms

Download the client

Travel copilot

Discover → compare → route → share

Search + Routes + Deep links

Follow the tutorial

Fleet operator

Inspect → diagnose → approve → route

InTouch + Search + Routes

Follow the tutorial

Agent governance

Purpose → allowlist → evaluate → release

6 patterns × 12 controls

Plan the agent