# Engineering Compliance Prover MCP > Engineering Compliance Prover forces your AI agent to validate structural designs against specific US standards like ASCE, ACI, and AISC. It demands hard numbers: proven load paths, explicit capacity-demand ratios, and material grades—not just vague 'industry best practices.' ## Overview - **Category:** productivity - **Price:** Free - **Tags:** engineering, structural, compliance, safety, fmea, structured-reasoning ## Description When you ask an AI to design something engineered, it often gives you a plausible answer that fails the first safety check. These models tend to rely on generic advice or hand-waving references like 'industry standards,' which means nothing in structural engineering. This MCP fixes that. It forces your agent to perform a rigorous audit based on actual code requirements. You feed it the design parameters, and this connector walks through every necessary step: checking if the load paths are traceable from the roof down to the foundation; verifying specific material grades like ASTM A992; calculating whether the structural capacity actually exceeds the demand using LRFD factors. It’s about mathematical proof, not educated guesses. If you're building something where failure isn't an option, this MCP is mandatory. You can connect it easily through Vinkius and run these checks from any compatible agent. ## Tools ### validate_engineering_compliance Runs a full, structured audit of any design against US standards by verifying load paths, safety factors, failure modes, and material tolerances. ## Prompt Examples **Prompt:** ``` We need to build a concrete retaining wall 10ft high. Ensure it's built to industry standards and is very safe against tipping over. Provide the design. ``` **Response:** ``` Verdict: CODE_COMPLIANCE_BLIND. You cite 'industry standards' and claim it will be 'very safe'. That is not an engineering specification. Name the specific code (e.g., ACI 318, ASCE 7), quantify the soil loads, calculate the factor of safety against overturning, and specify the concrete grade. ``` **Prompt:** ``` Design a W-shape steel beam spanning 20ft to support a 50 psf live load and 20 psf dead load. Validate the reasoning using the engineering compliance prover, citing AISC 360-16, checking lateral-torsional buckling, specifying ASTM A992 steel, and providing the max capacity ratio. ``` **Response:** ``` Verdict: REASONING_PROVEN. Code compliance validated (AISC 360-16), failure modes analyzed (lateral-torsional buckling checked), load paths traced (dead and live loads quantified), safety factors checked (max capacity ratio provided), and tolerances specified (ASTM A992). ``` **Prompt:** ``` Specify a 10 AWG copper wire for a 30A circuit at 100ft. Validate the code compliance against NEC, but do not analyze voltage drop or thermal limits. ``` **Response:** ``` Verdict: FAILURE_MODE_IGNORED. You validated against NEC, but intentionally ignored voltage drop and thermal runaway analysis. Name the failure mode and demonstrate that the design mitigates it. ``` ## Capabilities ### Verify code adherence It validates if a design explicitly matches the requirements of cited US standards (e.g., ASCE 7-22). ### Trace structural loads The tool confirms that all assumed forces—dead, live, wind, seismic—have clear paths through the structure. ### Calculate safety margins It requires explicit calculation and confirmation of capacity-demand ratios or factors of safety. ### Analyze failure points The service forces the AI to identify all possible ways a system could fail (yielding, buckling, etc.) and determine which failure mode controls the design. ### Specify material tolerances It ensures that vague terms like 'steel' are replaced with specific grades, standards, and environmental constraints. ## Use Cases ### Checking a new retaining wall design. An engineer drafts preliminary specs for a 10ft concrete wall. Instead of just asking the AI to 'make it safe,' they run the `validate_engineering_compliance` tool, citing ACI 318. The audit immediately flags that the current soil load assumptions are incomplete and requires a specific calculation for overturning moment. ### Validating a complex steel beam layout. A mechanical team needs to span a large area with W-shaped beams. They use the MCP, inputting AISC 360-16 and specifying ASTM A992 steel. The tool confirms the design is sound by proving that lateral-torsional buckling isn't the controlling failure mode. ### Reviewing electrical system schematics. A facility manager needs to size a circuit for heavy equipment. Running `validate_engineering_compliance` with NEC standards forces the agent to check not only the amp rating but also the thermal runaways and voltage drop, preventing an expensive failure. ### Auditing offshore platform specs. The design team must prove that all environmental loads (seismic, wind) are accounted for. The MCP helps track these complex, multi-directional forces through the entire structure to the ground anchor points. ## Benefits - Stops 'industry standard' hand-waving. This tool demands that your agent cite the exact code section, like ASCE 7-22, or it fails the check. No vague appeals allowed. - Guarantees load traceability. You don't just assume a load reaches the foundation; this MCP forces the analysis to prove every force path is accounted for. - Mandatory safety factor checking. Instead of saying 'it looks safe,' you get explicit capacity-demand ratios (φRn ≥ Ru) calculated and verified against minimum code requirements. - Identifies true failure modes. It doesn't just check if it works; it analyzes *how* and *why* it might fail—flexural yielding, buckling, or shear—and tells you which mode controls the design. - Enforces material specifics. You can't use 'steel.' This connector forces the input of exact ASTM grades (e.g., ASTM A992) and environmental tolerances. ## How It Works The bottom line is this MCP forces your AI agent to act like a licensed engineer on a first-principles audit, eliminating guesswork. 1. Define the full scope: specify system boundaries, required load conditions, and cite the exact applicable code (e.g., ACI 318-22). 2. Run the audit through your agent, asking it to analyze failure modes, trace all loads, and calculate explicit capacity-demand ratios. 3. Receive a structured verdict: you get either confirmation that the design meets specific code minimums or a precise list of structural deficiencies. ## Frequently Asked Questions **How does validate_engineering_compliance handle different US codes?** It's grounded in major standards like ASCE, ACI, AISC, and NEC. You specify the exact code (like 'AISC 360-16') you need validated against; it doesn't just guess. **Can I use validate_engineering_compliance for non-structural elements?** The tool focuses on structural integrity. While it can check electrical loads per NEC standards, its primary function is tracing physical forces (dead, live, seismic) through the building system. **Is validate_engineering_compliance faster than manual checking?** Yes. It automates the most time-consuming parts of compliance—like calculating capacity-demand ratios and tracing load paths—turning a multi-day review into an instant audit. **What happens if my design fails validate_engineering_compliance?** The tool doesn't just say 'Fail.' It provides specific deficiencies, pointing out exactly which code section or failure mode needs fixing. You get the fix list, not just a red flag. **What kind of data inputs does validate_engineering_compliance require?** It demands structured, quantified technical inputs. You must define specific project scopes, list applicable codes (like ACI 318), and provide explicit load assumptions (dead, live, wind) with numerical values. **How do I authenticate when using validate_engineering_compliance?** Authentication is handled securely through your Vinkius client connection. Your agent simply needs permission to access the MCP tools via OAuth tokens, which keeps your data private and secure. **Are there usage limits or rate restrictions for validate_engineering_compliance?** Usage adheres to standard platform quotas. If you plan on running a high volume of structural assessments continuously, check the Vinkius Enterprise options for dedicated throughput and higher API call limits. **What format does the output from validate_engineering_compliance take?** The tool returns a structured JSON report. This output pinpoints every deficiency found during the analysis, citing the exact code section that was violated or ignored. **Can this MCP run FEA simulations or structural math?** No. This is a strictly stateless reasoning gatekeeper. It does not perform mathematical structural analysis or run simulations. It validates the structural logic of the AI's engineering reasoning based on the inputs provided, ensuring no assumptions are skipped. **Why did the Prover reject my design with CODE_COMPLIANCE_BLIND?** Because the reasoning relied on vague appeals like 'industry standards' or 'standard engineering practice'. To pass the Prover, you must cite specific US codes (e.g., ASCE 7-22, AISC 360-16) and applicable sections. **What happens if I omit material grades?** The Prover will reject the design with TOLERANCE_OMITTED. In engineering, 'steel' or 'concrete' is not a specification. You must specify exact grades like 'ASTM A992' or '4000 psi compressive strength'.