SysML v2 systems engineering: generate, verify, simulate and export system models from text.
Scored across 32 tools
Most tools have distinct scope, but get_job_result and get_job_status carry identical descriptions, making them effectively indistinguishable at selection time. export_model versus mbe_export and build_full_system versus generate_system_model + build_3d_model + verify_model + run_simulation also invite misselection between the composite and the individual operations.
All names are snake_case with the verb first, which reads predictably. The only deviation is the mbe_ namespace prefix used for the collaboration/target family (mbe_submit_fragment, mbe_reconcile), which looks intentional rather than chaotic.
32 tools is heavy for a single server, and the surface splits into two large families (paid generation/verification jobs and the mbe_ collaboration workflow) plus job plumbing. Several tools could be consolidated, e.g. the job status/result pair and the composite build_full_system.
The lifecycle is well covered: generation, verification, simulation, trade study, safety analysis, export, 3D build, and a full fragment/reconcile/commit/baseline collaboration flow. However, there is no tool to list, fetch, or delete saved models, so agents lack an entry point to enumerate or clean up persisted m_ models.
Compile a saved systems model into placed 3D envelopes and existing authored procedural geometry. Returns a studio link and supported geometry formats. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Generate the system model, build its existing 3D geometry, verify requirements and allocations, and run selected declared simulations in one prepaid job. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. estimate_cost is optional and free, and with cost_mode "agent_authored" it validates a specification without running it. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Free. Stop the job and settle consumed work, refunding the remainder. The first line reports tokens and credits used; show it to the user.
Inputs: job_id (required).
Free. Discover MBEstudio systems engineering capabilities, enabled formats, pricing and the purchase URL. No model calls.
Inputs: no named parameters.
Free. Estimate the maximum model, compute and platform credits for a paid tool. Optional: paid tools can be called directly. Use it when the user asks what a job will cost. With cost_mode "agent_authored" and a specification, it also validates the specification with exactly the checks the paid tool runs and lists every problem as a blocker, so fix those before paying. No model calls or paid workload execution.
Inputs: tool (required), rows, model, study, +16 more.
Export a saved systems model using an advertised existing format: SysML, STL, scene JSON, engineering CSV, report HTML or existing result JSON. The result carries the file as content and as download_url, a signed link valid for 7 days that opens in any browser; give that link to the user. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Draft FHA failure conditions for a saved model's functions and FMEA failure modes for its parts under a chosen standard (ARP4761, DO-178C, ISO 26262, MIL-STD-882E or IEC 61508). Rows are written into a copy of the model as SysML, marked draft = true for an engineer to review; the saved model is not changed. With analysis "stpa" (no standard needed), drafts an STPA instead: losses, hazards, controllers with their process models, control actions and feedback as messages between parts, the unsafe control action table with all four types, and each safety constraint as a requirement; the result carries the compiler's STPA findings on the draft (missing UCA types, links to hazards and losses, unverified constraints, context variables no feedback updates). Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Generate a systems model or SysML architecture from a natural-language specification or structured agent-authored specification, including interfaces, components, requirements and constraints. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. estimate_cost is optional and free, and with cost_mode "agent_authored" it validates a specification without running it. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Free. Return this connection owner's prepaid balance and the existing purchase URL. No model calls.
Inputs: no named parameters.
Free. Read the prepaid job using the connection or key that paid for it. An exported file comes with download_url and filename; give the user the link rather than pasting a large file. The first line reports tokens and credits used; show it to the user.
Inputs: job_id (required).
Free. Read the prepaid job using the connection or key that paid for it. An exported file comes with download_url and filename; give the user the link rather than pasting a large file. The first line reports tokens and credits used; show it to the user.
Inputs: job_id (required).
Free. Register a SysML v2 API server and project as an export target. The token is stored encrypted and never returned.
Inputs: base_url (required), project_id (required), label, token, +1 more.
Free. Commit a proposal with every conflict resolved as one changeset on its branch, with stable ids and provenance; returns the changeset id. If the branch moved, the fragment is reconciled again first.
Inputs: proposal_id (required).
Free. Name a version of the model as a baseline that never moves.
Inputs: name (required), model_id (required), ref, message.
Free. Elements added, removed, changed and renamed between two commits, branches or baselines, matched by stable id.
Inputs: from (required), model_id (required), to.
Free. Compare a target's package with the model's current slice element by element: changed, only in the tool, or missing from it.
Inputs: target (required), model_id (required), subsystem_id (required).
Free. Export a subsystem and its dependencies (types, interface and item definitions, requirements, allocations; anything else as reference proxies) into a package of a target. dry_run returns the diff (create, update, unchanged, delete candidate) and touches nothing; create refuses a package that already holds it; update writes in place through the id map and never duplicates.
Inputs: target (required), model_id (required), subsystem_id (required), target_package (required), +3 more.
Free. Everything an agent needs before generating a subsystem: its boundary contracts (the ports, items, units and rates both sides of each interface state), the interfaces frozen during parallel generation, the functions allocated to it, the requirements on it, its current text, and the baseline id to generate against.
Inputs: model_id (required), subsystem_id (required), ref.
Free. Read a target package back (for file targets, from an uploaded file) as a fragment and reconcile it. Tool data never overwrites the model: it becomes a proposal like any agent's.
Inputs: target (required), model_id (required), file, package, +1 more.
Free. A proposal's conflicts grouped by type, with the resolutions each allows and any resolution already made.
Inputs: proposal_id (required).
Free. The package tree of a target, to choose target_package for mbe_export.
Inputs: target (required), model_id (required).
Free. Packages and package level definitions of a saved model that an agent can populate, with the agent or user that owns each.
Inputs: model_id (required), ref.
Free. Where subsystems can be exported: SysML v2 API servers this account registered (Ansys SAM, SysML v2 Cameo or CATIA Magic, the OMG Pilot), Cameo through XMI, and .sysml files, with what each can do.
Inputs: no named parameters.
Free. Merge a staged fragment into the current baseline in context and return the merge proposal: every change, and typed conflicts (identity, scope, ownership, frozen interface changes, port, item, unit and rate mismatches across boundaries, unallocated or doubly allocated functions, untraced requirements, SHACL, compile errors), each auto resolved by policy or waiting for review.
Inputs: fragment_id (required).
Free. Settle one conflict: accept the fragment's change, reject it (the element reverts to the baseline), approve an interface change, replace the element with corrected text, or for identity conflicts match a baseline element or declare it new.
Inputs: resolution (required), conflict_id (required), by, note.
Free. Stage a generated subsystem (SysML v2 text or element JSON) with its provenance. Staged only: it never changes the model. Then call mbe_reconcile.
Inputs: fragment (required), model_id (required), branch.
Ask the agent that owns a subsystem to try again, with your instruction. It sees the model, its last answer and the conflicts still open; its new answer is checked in context, retried on blocking conflicts, and replaces the open proposal for that subsystem. Nothing is committed. Needs model_id (a saved m_ model), subsystem and prompt. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Engineer a system with a swarm of agents (ADR 0027): an orchestrator divides the description into subsystems and frozen interface contracts, divides again any subsystem too large for one agent (max_depth levels, default 2), specialists write the subsystems in parallel, and every answer is checked in the context of the whole model, with conflicts sent back to the agent that caused them. With a prompt alone, the result is a new saved model holding the skeleton and one federation proposal per subsystem. With model_id (a saved m_ model), nothing is divided: one agent develops each subsystem the model already has, keeping what is there, and its work is staged as a proposal on that same model; pass subsystem to go one level deeper, decomposing that subsystem's parts, and repeat level by level. The prompt is then optional guidance. Nothing is written until you review and commit each proposal. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Run the saved model's declared state machines or an existing operational scenario. Reports transitions, requirement verdicts over time and unmatched events. No fluid, rigid-body or FMU solver. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Run an existing declared trade study over at most 256 design points and return objective values and the Pareto front. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Check a saved model with the native SysML compiler and SHACL shapes for requirement subjects and, when a structured specification exists, function allocations. Returns findings and explicit coverage. Paid. Call it directly: it charges what the job uses, never more than its estimated maximum or the max_spend_credits you set. request_id is optional; set one to make a retry safe. Returns job_id immediately; then call wait_for_job with it until finished is true.
Inputs: rows, model, study, events, +17 more.
Free. Wait for a job from a paid tool: returns as soon as it finishes (or, with until "changed", as soon as its status changes), else after wait_seconds (default 45, at most 50). Call it again while finished is false. A finished job comes with its result and studio_url; show the user the first line (tokens and credits) and the studio_url. Sends progress notifications while it waits when the client asks for them.
Inputs: job_id (required), until, wait_seconds.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Manage SysML2 projects; check, evaluate, verify, query, document and diagram models.
Build, run and analyse block-diagram simulations: PID, Bode, FFT, optimisation and code export.
Schema modeling in JSON, JSON-LD, and other formats with CoreModels platform.
Design domain models and generate deterministic multi-stack code, driven by your coding agent.
Add one secure layer between your agents and this server.