Jarvis
Provides optional local LLM support for analyze and free-form interpretation in the Jarvis engineering engine.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@JarvisDiseña un dron para carga de 2 kg con autonomía de 30 minutos."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Jarvis
v0.4.1
Deterministic engineering engine for designing physical systems with AI-assisted natural language.
Jarvis is aerial-first (drones and related vehicles): you describe goals and components in Spanish; calculation and simulation stay rule-based and auditable. The model may interpret — it does not invent the physics.
Read the one-page contract: VISION.md.
Quick start
git clone https://github.com/mcuestamunoz/Jarvis.git
cd Jarvis
python3 -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -e ".[dev,mcp]"CLI chat (optional: local Ollama for analyze / free-form interpret):
python -m jarvis.main --chat
# or
jarvis --chatPizarra (visor 3D + Continuity spatial assembly; huecos de arquitectura = slots, no BOM; mutación = CLI / Continuity / Situar drag):
jarvis boardTests:
pytestMCP server (Cursor / MCP clients):
python -m jarvis.adapters.mcp.serverWorkspace projects live under workspace/ (override with JARVIS_WORKSPACE_ROOT).
Ollama defaults: JARVIS_OLLAMA_BASE_URL, JARVIS_OLLAMA_MODEL (see src/jarvis/config.py).
Related MCP server: COMSOL MCP Server
What v0.4.1 includes
Everything in v0.4.0 (Continuity spatial assembly), plus Board Situar:
C-113 — Scene3D drag →
board_pose_bridge→ sameset_component_declared_box_posewriter as CLIdeclara…Free camera while situating (no forced cenital); screen-plane drag follows the cursor; Shift = profundidad (Y)
Situar UX: larger pane, zoom range, live preview
Standoff count gate B4-min (
count==4→ corners; else omit)
Live suite 2669 · UI 80
Locks unchanged: Prop/Energy = HD-004 wall; System Optimization deferred until pain; screening AABB ≠ fit VERIFIED
Next (not in this tag)
Fit VERIFIED (PRIORIDAD) — beyond
"cabe"screeningIDLE frame-part count (
6 standoffs) · N≠4 standoff layout · sourced dims #4 · LLM pending deactivate
See docs/IMPLEMENTATION_TASKS.md.
Docs
Doc | Role |
Product contract (non-technical) | |
v1 usable — when a project is “done” | |
A' — Situation / Evidence / Next useful step | |
How the system is built | |
System Map — connections & authority | |
Roadmap, gaps, software/product debt | |
Physics debt gated on T1/T2 lab (ESC η, battery C-rate, sag, OP→consumo) | |
Living CLI findings register (G9–G20) | |
Deeper product / flow reference |
Tags
v0.4.1 / checkpoint-board-situar — Board Situar drag→pose (C-113) + free camera + standoff count gate; suite 2669 · UI 80.v0.4.0 / checkpoint-continuity-spatial-assembly — Continuity spatial assembly (situar el mapa); suite 2652.v0.3.8 / checkpoint-spatial-board-projector — ship spatial_board.py (was gitignored); /workspace/ ignore.v0.3.7 / checkpoint-structure-representation-closed — Structure representation arc closed (catalog→parts→rebind→plates); suite 2294.v0.3.6 / checkpoint-experimental-prop-energy-closed — experimental prop/energy/Structure A/fail-routing construction closed; knowledge-parity phase starts.v0.3.5 / checkpoint-phase25-hover-energy — Phase 2.5 honest hover-regime autonomy.v0.3.4 / checkpoint-motor-op-voltage-coherence — motor OP voltage gate + DSE live params.v0.3.3 / checkpoint-validation-case-regression-gate — Validation Case probe/docs.v0.3.2 / checkpoint-deferred-queue-cd — Deferred Queue C+D.v0.3.1 / checkpoint-next-engineering-block — G24-A + P2-2 OP bridge.v0.3.0 / checkpoint-propeller-catalog-bind — propeller help-choose → exact OP.v0.2.0 — H1–H4 handoffs closed; System Map at 0 RED.v0.1.0-prototype — first functional cut.
This server cannot be deployed
Maintenance
Related MCP Connectors
Official DevSpeak MCP server — translate technical text into formal specs from any AI IDE or agent
MCP server for aerospace calculations: orbital mechanics, ephemeris, DSN operations, ...
- mcpOAuthcom.gibsonai
GibsonAI MCP server: manage your databases with natural language
MCP server for generating rough-draft project plans from natural-language prompts.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceMCP server for creating, modifying, and analyzing KiCAD schematic files using natural language.20MIT
- AlicenseBqualityDmaintenanceMCP Server for COMSOL Multiphysics simulation automation via AI agents.781MIT
- AlicenseNot gradedqualityBmaintenanceA local MCP server that enables AI agents to create and edit parametric CAD models through natural language, using a validated operation graph that compiles to real geometry.1MIT
- AlicenseAqualityBmaintenanceMCP server providing embedded engineering calculators and code generators as tools for AI agents, enabling precise, deterministic embedded math and C code generation.2938 npmMIT