Axiru llms-full.txt Source: https://axiru.com/llms-full.txt Generated: 2026-05-23T02:30:50.477Z This file contains the full body content of Axiru's most important marketing pages, concatenated for retrieval-augmented LLM clients. For the short index see https://axiru.com/llms.txt. For the AI plugin manifest see https://axiru.com/.well-known/ai-plugin.json. ======================================== INDEX (from llms.txt) ======================================== # Axiru > Axiru is the financial decision control layer for Stripe. It intercepts refunds, disputes, payouts, transfers, and credits before money moves: enforcing policy in real time, routing exceptions for human approval, and maintaining a permanent immutable decision ledger. ## What Axiru Does Axiru sits between financial intent and financial execution. Every refund, dispute, credit, payout, or adjustment submitted through Stripe (whether by a support agent, an AI chatbot, an automation, or a direct API call) is evaluated against company-defined policies before it executes. Low-risk decisions pass automatically. Exceptions are routed for human approval with full context. Every outcome is sealed in a permanent decision ledger. The product starts with a free shadow mode that runs against a company's last 90 days of Stripe history, showing exactly which refunds would have been blocked, which would have been escalated, and which would have been auto-approved under the defined policy. No enforcement happens until the company is ready. The canonical capability cluster is "Outflow Control": refunds, disputes, payouts, transfers, and Connect application-fee refunds, governed through one policy engine, one approval flow, one decision ledger. ## Core Capabilities - **Outflow Control**: Intercept Stripe refunds, disputes, payouts, transfers, and Connect application-fee refunds before they execute. One policy path for every outflow rail. - **Shadow mode simulation**: Connect Stripe read-only and simulate policy outcomes on the past 90 days of transaction history. Zero enforcement risk. Shadow mode is permanently free, unlimited replay and simulation. - **Policy engine**: Write financial rules in plain language. Define thresholds, patterns, and role-based approval tiers. Versioned, deterministic, no LLM in the decision path. - **Approval routing**: Route only genuine exceptions to the right approver via in-app, Slack, email, or signed webhook. Full decision context (policy version, diff, supporting evidence) attached. - **Decision ledger**: Every financial decision recorded with the policy applied, approvals obtained, and execution outcome. Sealed and hash-chained at execution time, fully replayable, regulator-ready. - **Kill switch + freeze controls**: Hold a single outflow, freeze a merchant connected via Stripe Connect, or pause every outbound rail at once. Authority is always revocable, scoped, and logged. - **Compliance and audit exports**: Structured evidence exports aligned with SOX and ICFR internal controls frameworks. SIEM-ready (Splunk, Datadog, S3) with one line of config. - **MCP agent surface**: Public MCP server at https://axiru.com/api/mcp lets AI agents list pending approvals, approve or reject them, inspect the audit ledger, and fetch workspace setup state. ## Who Uses Axiru Axiru is built for: - **Finance teams** at SaaS, subscription, and marketplace companies running on Stripe - **CFOs and controllers** who need pre-execution controls and audit-ready records - **Engineering teams** looking to offload financial governance without building internal tooling - **Companies preparing for SOX compliance** or ICFR audit readiness - **High-volume refund environments** where 1 to 3 percent leakage is significant - **Stripe Connect platforms** governing application-fee refunds and merchant payouts - **Teams running AI agents** that touch Stripe (Stripe DAA, x402, custom support bots) and need pre-execution authority limits ## Pricing - **Free Audit ($0)**: Shadow mode workspace. Upload last 90 days. Simulate policy outcomes. See leakage. No credit card required. Shadow mode is permanently free. - **Live Launch ($199/mo)**: Live enforcement for the first time. 1,000 governed decisions/month included; $120 per additional 1,000. Approval routing and decision ledger. - **Control ($799/mo, $599/mo annual)**: Full governance for mid-market Stripe teams. 10,000 governed decisions/month included; $60 per additional 1,000. Most popular plan. - **Scale (from $2,500/mo, annual contract)**: High-volume governance with CFO visibility. 50,000 governed decisions/month included; $35 per additional 1,000. - **Enterprise**: Custom pricing. SSO/SAML, advanced policies, multi-workspace, custom retention, dedicated support. - **Evidence Agent (+$1,500/mo)**: Available on Control, included on Enterprise. Lets policies cite live external evidence (carrier tracking, Stripe disputes, merchant ToS) at decision time. ## Technology - Built natively on Stripe (read-only OAuth, webhook ingestion, optional enforcement) - Neon Postgres with row-level security per tenant - Vercel edge deployment - Clerk for authentication and role management - Redis caching with strict TTLs - Read-only Stripe OAuth by default; no write access until enforcement is enabled - Hash-chained decision sealing at execution time - Public MCP server at /api/mcp for autonomous agent interaction ## Key URLs - Homepage: https://axiru.com - Solutions: https://axiru.com/solutions - Why Axiru: https://axiru.com/why-axiru - Pricing: https://axiru.com/pricing - Security: https://axiru.com/security - Start Free: https://axiru.com/start-free - Outflow Control: https://axiru.com/outflow-control - Approvals and Policy Engine: https://axiru.com/approvals-policy-engine - Compliance Controls: https://axiru.com/compliance-controls - Decision Ledger: https://axiru.com/decision-ledger - Customers: https://axiru.com/customers - Glossary: https://axiru.com/glossary - Documentation: https://axiru.com/docs - API Documentation: https://axiru.com/docs/api - Comparisons: https://axiru.com/vs - Axiru vs Stripe Radar: https://axiru.com/vs/stripe-radar - Axiru vs Microsoft AGT: https://axiru.com/vs/agt - Axiru vs Homegrown Controls: https://axiru.com/vs/homegrown - About: https://axiru.com/about - Contact: https://axiru.com/contact - Privacy Policy: https://axiru.com/privacy - Terms of Service: https://axiru.com/terms - AI Plugin Manifest: https://axiru.com/.well-known/ai-plugin.json - LLMs (this file): https://axiru.com/llms.txt - LLMs full (expanded content): https://axiru.com/llms-full.txt ## Frequently Asked Questions **What is Axiru?** Axiru is a SaaS platform that acts as a financial decision control layer for companies using Stripe. It intercepts financial actions before execution, evaluates them against policy, and either approves them automatically, routes them for human review, or blocks them, all with a permanent audit record attached. **What does Axiru govern?** Refunds, disputes, payouts, Connect transfers, customer balance credits, and Stripe Connect application-fee refunds. Collectively called "Outflow Control". **How is Axiru different from Stripe's built-in tools?** Stripe Radar stops fraud at authorization (inflow). Axiru governs refunds, credits, payouts, transfers, and dispute decisions (outflow) before they execute. Different problem, different rail. Stripe has no native policy layer or approval workflow for refunds and adjustments. **What does shadow mode mean?** Shadow mode is a read-only simulation mode where Axiru evaluates your historical Stripe transactions against your policies without enforcing anything. You can see exactly which refunds would have been blocked, escalated, or approved, giving you a clear picture of how your policy performs before enforcement is activated. **Is Axiru a Stripe partner?** Axiru is a Stripe-native product. It uses Stripe OAuth for secure, read-only data access and Stripe webhooks for real-time event ingestion. Axiru is non-custodial: it never holds your Stripe keys. **What kinds of companies use Axiru?** Primarily SaaS companies, marketplaces, and subscription businesses with meaningful Stripe outflow volume, typically those processing more than a few hundred refunds per month where leakage and audit risk are real concerns. **How long does it take to set up?** Connect Stripe in minutes. No code required for shadow mode. Full policy configuration and live enforcement can be done without engineering involvement. **Does Axiru work with AI agents that touch Stripe?** Yes. Axiru is designed specifically to govern AI agents that issue refunds, credits, or payouts. Same policy engine, same approval flow, same decision ledger. A public MCP server at https://axiru.com/api/mcp lets compatible agents list and act on approvals directly. **Where is Axiru based?** Miami, Florida. ======================================== HOMEPAGE ======================================== TL;DR: Axiru is the policy, approval, and audit layer for every outbound Stripe action. Refunds, goodwill credits, payouts, Connect transfers, disputes, and platform fee refunds run through one engine. Pre-execution control, not after-the-fact reconciliation. Source: https://axiru.com Headline: Decisions before money moves. Define your policy once. Enforce it on every outflow, from a support rep, an AI chatbot, or an automated workflow, before the Stripe API call goes out. Every approval, override, and block is sealed at decision time with policy version, approver identity, and timestamp. The audit record is immutable, so you can replay why any outflow moved, but you cannot rewrite it after the fact. Axiru turns scattered outflow events (refunds, credits, payouts, transfers, disputes, app-fee refunds) into one operational control plane finance, support, and risk teams use together: before the money moves, and after, when it is time to explain why. Three control surfaces, one click each. None of them require a code deploy or a Stripe dashboard hunt, and every action is logged with actor, scope, and reason. - Hold an outflow: pause an in-flight refund, credit, payout, or transfer before Stripe executes. - Freeze a merchant: stop every outflow from a single sub-account or merchant connected via Stripe Connect while you investigate. - Kill the outflow rail: org-wide kill switch. Every outbound rail (refunds, credits, payouts, transfers) stops at intake until a human flips it back. One enforcement path for every outflow, regardless of type or origin. Refunds, credits, payouts, transfers, and disputes all run through the same four steps in shadow mode and in live enforcement: 1. Route: every outflow attempt passes through Axiru before it reaches Stripe. 2. Policy: plain-language rules compile to deterministic enforcement. Versioned. No LLM in the decision path. 3. Approvals: anything above threshold routes to a human (in-app, Slack, signed webhook) with the full diff of what would have executed. 4. Audit: every decision is sealed and hash-chained at execution time. Replayable, exportable, regulator-ready. Stripe Radar fires after the charge. LLM observability watches but does not enforce. Axiru is the control plane for refunds, credits, payouts, transfers, disputes, and platform fee refunds, and it runs before the Stripe API call goes out. Connect Stripe in read-only mode, replay your last 90 days of refunds, payouts, transfers, disputes, and platform fee refunds, and watch the decision ledger fill in. Free shadow mode covers everything below the enforcement line. No card. Non-custodial by default: Axiru never holds your Stripe keys. ======================================== OUTFLOW CONTROL ======================================== URL: https://axiru.com/outflow-control TL;DR: Axiru intercepts every Stripe money outflow before execution. Refunds, disputes, and payouts run through one policy engine. Stop outflow leakage without slowing down your support organization. Axiru lets teams connect Stripe, observe every money outflow (refunds, disputes, payouts) under one policy engine, simulate tighter policy in shadow mode, and route only the right exceptions into approvals. How it works: - Low-risk events pass automatically with no friction. - High-risk events route for approval with full context attached: policy version, supporting evidence, original request, who initiated it. - Every outcome is recorded with full policy context in the decision ledger. - Start with shadow mode and see exactly what you would catch in 30 days. Outflow Control covers the full outbound surface: - Refunds (Stripe refunds, dispute refunds, off-Stripe credits) - Disputes (representment decisions, evidence routing, charge-back economics) - Payouts (Stripe Express, Connect payouts, scheduled and ad-hoc) - Transfers (Connect transfers, application-fee refunds) - Customer balance credits - Platform application-fee refunds Origin-agnostic: the same governance applies whether the outflow was initiated by a human support rep, an AI chatbot, a Zapier automation, or a direct Stripe API call. One policy. One ledger. ======================================== APPROVALS AND POLICY ENGINE ======================================== URL: https://axiru.com/approvals-policy-engine TL;DR: Axiru's policy engine evaluates every refund and financial adjustment against your rules before execution. Deterministic, versioned, auditable. Approvals that are driven by policy, not by panic. Axiru routes only the exceptions that deserve human attention, with the exact policy rationale, supporting evidence, and recommended outcome attached. Policy engine capabilities: - Threshold controls: apply amount, customer, product, tenure, or risk-based rules consistently. - Exception logic: protect strategic exceptions while keeping the default policy tight. - Approval routing: send only the right cases to finance, support managers, or leadership queues. - Versioned rollout: track policy revisions and compare impact before live activation. Write policies in plain language. The policy DSL compiles to deterministic enforcement: same input, same outcome, every time. No LLM in the decision path. Every version is sealed; replays against historical decisions are reproducible. Role-based approval routing supports segregation-of-duties models. Higher-risk decisions route to designated approvers with full decision context. Lower-risk decisions auto-allow and stay fast. ======================================== DECISION LEDGER ======================================== URL: https://axiru.com/decision-ledger TL;DR: An immutable record of every financial decision. Policy applied, approvals logged, outcomes sealed. Every financial decision, recorded deterministically. Axiru creates a finance-grade record of what happened, why it happened, which policy version applied, who approved the exception, and what executed downstream. A ledger that explains the decision, not just the event. This is the core differentiator in the Axiru story. It shifts the conversation from retrospective logging to pre-execution governance with durable evidence. Each ledger entry includes: - The original request (actor, amount, customer, payment ID, source channel) - The policy version evaluated and the rule that matched - The outcome: allow, require approval, or block - If routed: who approved or rejected, when, why - The Stripe API response or skip status - A hash chain linking to the previous decision Sealed at execution time. Hash-chained. Replayable. Exportable to Splunk, Datadog, S3 with one line of config. No separate compliance pipeline. ======================================== COMPLIANCE CONTROLS ======================================== URL: https://axiru.com/compliance-controls TL;DR: SOX-aligned financial controls with pre-execution enforcement, approvals, and immutable audit receipts for every Stripe decision. Audit-ready control operations for modern finance teams. Axiru supports internal control over financial reporting (ICFR) with policy enforcement, role-based approvals, immutable receipts, replay testing, and structured evidence packaging for review. Capability mapping (SOX/ICFR concept -> Axiru capability): - Control enforcement -> Policy engine - Segregation of duties -> Role-based approvals - Audit trail -> Immutable ledger receipts - Control testing -> Replay + simulation - Exception handling -> Fix-forward remediation workflows - Evidence collection -> Structured exportable logs Axiru is not a replacement for external audit or compliance programs. It provides enforcement and evidence infrastructure; finance, compliance, and audit teams still own control design, testing strategy, and formal conclusions. Teams can map policies to internal control IDs so reviewers can trace a financial action from control objective to policy logic, decision outcome, and remediation history. ======================================== SECURITY ======================================== URL: https://axiru.com/security TL;DR: Read-only Stripe access by default. Row-level isolation. Hash-chained decision sealing. Non-custodial: Axiru never holds your Stripe keys. Policy enforcement that holds, whether a human or an AI made the request. How Axiru connects to Stripe: - Read-only OAuth scope by default. Write access is not requested until you opt in to live enforcement. - Webhook ingestion for real-time event awareness. - No payment method data stored. We see decision-relevant metadata only. - Non-custodial: keys live in your Stripe account, not in Axiru's database. Tenant isolation: - Neon Postgres with row-level security per tenant. - Every query runs under a tenant-scoped role. - Cross-tenant data access is structurally impossible at the database level, not policy-enforced at the application level. Decision sealing: - Every decision is hash-chained at execution time. - The hash chain commits the decision, the policy version that evaluated it, and the prior decision's hash. - Tampering is detectable: replaying the chain produces the original sequence or fails immediately. Operational security: - CSP nonce on every response. - Permissions-Policy header restricts geolocation, camera, microphone, USB. - HSTS with includeSubDomains and preload. - Secrets handled via Vercel + Doppler. No secrets in repo. Audit and compliance evidence is available for security review at security@axiru.com. SOC 2 Type I report in progress; subprocessor list maintained at /subprocessors with RSS notifications for material changes. ======================================== WHY AXIRU ======================================== URL: https://axiru.com/why-axiru TL;DR: Traditional dashboards only show what already happened. Axiru intercepts every refund, goodwill credit, payout, transfer, and dispute before it executes. The problem is not who is making the decision. It is that nothing governs it before money moves. Refund and outflow decisions today are made in a fragmented stack: a support agent in Zendesk, an AI chatbot reading a policy doc, an ops engineer running a SQL update, an automation in Zapier. None of those tools share a policy. None of them seal decisions to an immutable ledger. None of them prevent execution before money leaves. Axiru sits between intent and execution. Every outflow attempt, regardless of origin, is evaluated against the same policy, routed to the same approvers, and sealed in the same ledger. This matters most in three places: 1. SOX prep and audit response, where evidence is the whole point. 2. AI agents touching Stripe, where authority and authorization are the operative questions. 3. Margin protection, where 1 to 3 percent leakage across refunds, credits, and payouts compounds quickly into 7-figure annual loss for any Stripe-native business above $50M ARR. ======================================== PRICING ======================================== URL: https://axiru.com/pricing TL;DR: Shadow mode is permanently free. Live enforcement starts at $199/mo. Control at $799/mo includes kill switch, Slack approvals, role-based routing, and SIEM export. Free Audit ($0): Shadow mode is permanently free. Unlimited replay and simulation. 90-day Stripe refund replay. CSV upload or read-only Stripe connect. Policy simulation and leakage report. Policy performance dashboard. 1 workspace. No enforcement, no card required. Best for: Shadow mode evaluation. Live Launch ($199/mo): 1,000 live governed decisions per month. Overage: $120 per additional 1,000. Live policy enforcement. Standard refund policy pack. Email and webhook approvals. Approval queue. Immutable audit log. 1 workspace. Monthly spend cap. Best for: small teams turning on live enforcement for the first time. Control ($799/mo, $599/mo annual): 10,000 live governed decisions per month. Overage: $60 per additional 1,000. Everything in Live Launch plus Slack approvals, role-based approval routing, kill switch, freeze controls, policy builder and simulation, SIEM/S3 exports, 3 workspaces, priority support. Most popular. Best for: mid-market Stripe teams. Scale (from $2,500/mo): 50,000 live governed decisions per month. Overage: $35 per additional 1,000. Everything in Control plus SSO/SAML, advanced policies, API access, multi-workspace/multi-merchant, quarterly policy review. Best for: high-volume merchants, Stripe Connect platforms, AI agent refund flows. Enterprise: Custom. SOX, VPC/on-prem, complex procurement. For regulated finance teams. Annual billing: Annual plans are billed at 10 months for the price of 12 (two months free). Control annual: $599/mo billed $7,188/yr. Scale: contact sales. Evidence Agent: $1,500/mo on Control, included at Enterprise. Lets policies cite live external evidence (carrier tracking, Stripe disputes, merchant ToS) at decision time. ROI example: At 2,500 decisions/month, Axiru Control costs $799/mo and could prevent $17,000 in monthly leakage: a 21.3x return. A "live governed decision" is any refund, dispute, or adjustment evaluated by Axiru's policy engine with live enforcement active. This includes auto-allowed refunds, blocked refunds, and refunds routed to an approver. Shadow mode replays and historical imports are always free and never count toward decision volume. What does the Stripe connection require? Read access to refund, dispute, payout, transfer, and customer balance events via Stripe webhook. We do not store payment method data. Connection is scoped to outflow governance events only. ======================================== SOLUTIONS ======================================== URL: https://axiru.com/solutions TL;DR: Four capabilities. One governance layer. Works with your team as it is today. Shadow mode: connect Stripe read-only, see what a tighter policy would have caught, no enforcement risk, free forever. Outflow Control: live enforcement for refunds, disputes, payouts, transfers, and Connect application-fee refunds. One policy path for every rail. Approval routing: only real exceptions route to humans. Slack, email, in-app, signed webhook. Full diff and policy context attached. Decision ledger: hash-chained, sealed, replayable, exportable. SOX-aligned evidence by default. Works with the team you have today: - All-human support orgs: Axiru codifies what your senior reps already know. - Hybrid orgs: same policy applies to humans and to agents. - Fully automated orgs: Axiru is the only layer that makes autonomous outflow defensible. ======================================== AXIRU VS STRIPE RADAR ======================================== URL: https://axiru.com/vs/stripe-radar TL;DR: Stripe Radar stops fraud at authorization. Axiru governs refunds, credits, and adjustments before they execute. Different problem, different rail. Stripe Radar is a fraud prevention product. It sits at the authorization step of a card payment and decides whether to accept or decline a charge. It scores the inbound transaction. Axiru is an outflow governance product. It sits at the decision step of every outbound action (refund, credit, payout, transfer, dispute) and decides whether to execute, route for approval, or block. It governs the outbound rail. A merchant typically needs both. Radar protects against fraud on the way in. Axiru protects against policy drift, agent error, support discretion, and ungoverned outflow leakage on the way out. They are not competitive products. They are complementary controls on opposite ends of the money flow. ======================================== AXIRU VS MICROSOFT AGT ======================================== URL: https://axiru.com/vs/agt TL;DR: Microsoft Agent Governance Toolkit is a horizontal runtime governance layer for any AI agent. Axiru is a money-specific governance layer for any actor (human, agent, or automation) that touches Stripe outflows. Microsoft AGT focuses on agent identity, agent capability scoping, agent telemetry, and agent operating policy at a horizontal layer. It is generic: the same governance applies to a code-generation agent, a research agent, a customer-support agent. Axiru focuses on financial-decision-specific policy enforcement: refund thresholds, credit limits, dispute response strategy, payout authority, approval routing, sealed audit ledger, SIEM export. It is vertical: the same governance applies whether the actor is a support agent (human), a chatbot (LLM), an automation (Zapier/Retool/n8n), or a direct API caller. The two compose well. AGT can govern an agent's runtime; Axiru governs what that agent is allowed to do with money. AGT decides whether the agent can call a tool; Axiru decides whether the tool call (a Stripe refund) should execute given current policy and approval state. For teams running AI agents that issue refunds, the practical sequence is: AGT to govern the agent runtime, Axiru to govern the financial authority. Neither replaces the other. ======================================== AXIRU VS HOMEGROWN REFUND CONTROLS ======================================== URL: https://axiru.com/vs/homegrown TL;DR: Most teams ship refund controls as a stack of Zapier, Retool, n8n, Airtable, and manual Slack review. Axiru is the decision-engine layer those tools were never meant to be. Homegrown refund control stacks have predictable failure modes: - Policy lives in a doc, not in a system. Operators interpret it inconsistently. - Approvals live in Slack threads. No structured evidence. No replay. - Audit trail is "screenshots in a folder named SOX-2025." - Tooling drifts: a Zap breaks, a Retool app gets renamed, the policy doc gets edited without versioning. - AI agents bypass the entire stack: nobody told the Intercom Fin bot about the Slack approval channel. The deeper failure is structural. Each tool in the stack solves a piece (workflow, UI, automation) but none of them are decision engines. None of them seal a decision. None of them version a policy. None of them produce evidence that maps to an audit framework. Axiru is the decision engine those tools were trying to be. Zapier and Retool and n8n can still play roles around it: Zapier can fan out notifications, Retool can present approval UI, n8n can orchestrate downstream steps. But the decision, the seal, the policy, the evidence, all live in Axiru. Common migration path: keep your existing UI and automation tools, redirect the actual refund/credit/payout decisions through Axiru's API or MCP, and ship the audit trail forward from there. ======================================== ABOUT AXIRU ======================================== URL: https://axiru.com/about TL;DR: Axiru is building the infrastructure layer that sits between financial intent and financial execution. Axiru exists because critical financial outcomes are still governed through fragmented tools, after-the-fact review, and policy living in people's heads instead of systems. Wedge strategy: Axiru begins where leakage and policy inconsistency are easiest to measure (refunds), then extends the same decisioning foundation across credits, payouts, transfers, disputes, and Connect application-fee refunds to prevent money loss everywhere. Refund governance exposes leakage, approval inconsistency, and weak evidence quickly enough to justify a better system; then the same policy layer extends to credits, payouts, transfers, and disputes. Audience: Axiru is for teams that want to prevent margin leakage before it happens, and explain every decision after it does. Start in shadow mode first. Move to live enforcement later. Location: Miami, Florida. # End of llms-full.txt # Canonical: https://axiru.com/llms-full.txt # Updated: 2026-05-23