# Accounting & Audit Prover MCP > The Accounting & Audit Prover MCP forces your AI agent to validate financial arguments against specific US standards like FASB ASC and PCAOB rules. It demands quantified materiality thresholds, formal Risk of Material Misstatement assessments, and sufficient documentary evidence before accepting a conclusion. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** accounting, audit, gaap, fasb, pcaob, structured-reasoning ## Description AI agents can write professional-looking audit plans, but they often fail on the fundamentals of regulatory compliance. They might reference 'GAAP' generally without citing the exact FASB ASC or PCAOB section needed for validation. This MCP changes that. It forces your agent to stop guessing and start proving. You get a structured review process that demands specific dollar thresholds—not just vague concepts like 'significant amount.' The system requires you to quantify materiality, assess internal controls to determine risk, and list the exact evidence (like bank confirmations or signed contracts) backing every claim. This isn't about drafting memos; it's about building an auditable paper trail that stands up to scrutiny. Users connect this MCP through Vinkius, linking it directly into their existing AI workflows so they can run these rigorous checks without leaving their primary coding environment. ## Tools ### validate_accounting_audit Validates accounting or audit reasoning by forcing the agent to cite US standards, quantify materiality, assess risk of material misstatement (ROMM), and specify evidence. ## Prompt Examples **Prompt:** ``` A client recognized $50M in revenue before delivering the software. According to GAAP, this is wrong. They should wait until delivery. Validate this using the prover. ``` **Response:** ``` Verdict: STANDARD_COMPLIANCE_BLIND. 'According to GAAP' is not a citation. Cite FASB ASC 606, identify the specific performance obligation, quantify materiality, and specify the audit evidence required to verify delivery. ``` **Prompt:** ``` Analyze the revenue recognition of a $5M software contract under ASC 606. Validate the conclusion with the prover, quantifying materiality at $100k, assessing ROMM as High due to complex performance obligations, and detailing specific vouching procedures. ``` **Response:** ``` Verdict: REASONING_PROVEN. Standard compliance validated (ASC 606), materiality quantified ($100k), ROMM explicitly assessed (High), substantive procedures detailed, and grounded in evidence. ``` **Prompt:** ``` Under PCAOB AS 2315, we will audit the inventory balance. The materiality is $500k. We will perform tests of details on the warehouse counts. Validate this plan. ``` **Response:** ``` Verdict: ROMM_IGNORED. You cited the standard, set materiality, and proposed procedures, but completely skipped the Risk of Material Misstatement assessment. Is inherent risk high? Control risk low? Assess ROMM before testing. ``` ## Capabilities ### Validate US Standard Compliance Checks if an accounting conclusion cites a specific FASB ASC or PCAOB standard reference number. ### Quantify Financial Materiality Requires the agent to establish and apply specific dollar thresholds or percentage benchmarks when assessing misstatements. ### Assess Audit Risk Profile Mandates that the agent first assess the inherent risk and control risk before designing any audit procedures. ### Specify Substantive Procedures Forces detailed naming of specific tests, such as vouching to invoices or performing cut-off testing. ### Ground Conclusions in Evidence Ensures the final conclusion relies on citing sufficient, appropriate documentary proof (e.g., board minutes). ## Use Cases ### Reviewing complex revenue recognition A finance team member has an AI draft arguing for revenue recognition on multi-year SaaS contracts. Running the validation tool forces the agent to cite FASB ASC 606, quantify performance obligations, and detail which specific contract documents prove delivery. ### Planning a major inventory audit An auditor needs to validate their plan for physical count testing. The MCP demands that the agent first assess internal control risks (Control Risk) before it allows the proposal of tests, ensuring no procedural step is missed. ### Analyzing historical misstatements A compliance officer wants to know if a prior period error was properly handled. The tool forces quantification of materiality against pre-tax income and requires citing external legal counsel letters as evidence. ## Benefits - Stops vague 'GAAP' citations. If your agent suggests something, this ensures it references the specific FASB ASC or PCAOB section number needed for compliance. - Eliminates guesswork on money size. You force quantification of materiality, moving beyond general statements like 'significant amount' to hard dollar figures. - Builds risk awareness into every step. Before proposing any test, the system requires an assessment of the Risk of Material Misstatement (ROMM). - Ensures defensible procedures. Instead of recommending generic testing, you get specific instructions, like confirming balances with banks or vouching to invoices. - Closes the evidence gap. The output demands that every conclusion is tied back to a required documentary source, making the audit trail airtight. ## How It Works The bottom line is that you get back an objective assessment of whether your accounting argument meets professional audit rigor. 1. Define the precise scope of the audit or accounting issue, citing a specific financial statement area and an applicable US standard. 2. Run the validation tool, which will systematically check for quantified materiality, assessed risk (ROMM), and detailed procedures. It rejects any output missing these structural components. 3. Receive a verdict on your agent's reasoning: either 'REASONING_PROVEN,' meaning all standards were met, or a specific error code detailing the deficiency. ## Frequently Asked Questions **Can this MCP perform financial reconciliations?** No. This is a strictly stateless reasoning gatekeeper. It does not perform mathematical reconciliations, access ERP systems, or review ledgers. It validates the logical structure of the AI's accounting/audit reasoning based on the inputs provided. **Why did the Prover reject my accounting memo with STANDARD_COMPLIANCE_BLIND?** Because the reasoning relied on vague appeals like 'according to GAAP' or 'standard accounting practice'. To pass the Prover, you must cite specific US standards (e.g., FASB ASC 606, PCAOB AS 2110). **What does ROMM mean and why is it required?** ROMM stands for Risk of Material Misstatement. In auditing, you cannot propose substantive procedures without first assessing the inherent and control risks that could lead to a material error. The Prover forces this assessment. **How do I connect and use the `validate_accounting_audit` tool with my AI client?** You connect via Vinkius using any MCP-compatible client (like Cursor or Claude). The platform handles the secure connection, letting your agent access all validation tools without needing complex setup. **For `validate_accounting_audit`, what is the minimum required specificity for an 'AUDIT SCOPE'?** The scope must be highly detailed. Instead of naming a broad account like 'Revenue,' you must specify the exact area, such as 'ASC 606 revenue recognition for multi-year contracts.' Specificity is mandatory. **If my audit plan fails multiple checks (e.g., Materiality and ROMM) using `validate_accounting_audit`, how does the output show all deficiencies?** The tool's output will list every structural deficiency found, flagging them individually by their failure type (e.g., 'MATERIALITY_UNQUANTIFIED'). You must address each required pivot before proceeding. **Does `validate_accounting_audit` accept standards other than US GAAP/FASB ASC?** No, this MCP is strictly built for U.S. accounting and audit frameworks. To run the validation, you must cite explicit FASB ASC or PCAOB reference numbers. **Are there any usage limits or rate limits when running `validate_accounting_audit` repeatedly?** Vinkius manages resource access across its catalog. For optimal performance, we recommend grouping related audit areas and calling the tool in logical batches rather than rapid succession.