Adyen vs Paddle

Agent access

1. ai-native userPoint an agent at llms.txt or agent-oriented docs

   weight 2 · round to Adyen
   Adyen✓fullprobed10/10

   Adyen has a live, probed llms.txt (HTTP 200) plus markdown-renderable docs pages (append .md), a documented full-docs ZIP mirror explicitly for feeding AI agents/code assistants, and an OpenAPI spec repo positioned as LLM context — this is direct, independently-probed confirmation of agent-oriented docs.

   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.adyen.com/llms.txt # Adyen Docs > Developer and merchant documentation for Adyen payments, Adyen f…”
   • [probe] “PROBE docs-md: HTTP 200 at https://docs.adyen.com/online-payments.md --- title: "Online payments" description: "Accept payments with cards, …”
   • [claimed-docs] “"Adyen makes documentation accessible to developers as well as the software and AI tools they create." Documented agent-docs surface: llms.t…”
   • [claimed-docs] “The Adyen OpenAPI Specification repository (github.com/Adyen/adyen-openapi) "contains OpenAPI specifications for every public API Adyen offe…”

   proof ↗⚑ flag

   Paddle✓fullprobed9/10

   Paddle publishes a dedicated llms.txt (confirmed live via HTTP 200 probe) plus a full suite of agent-oriented docs including agent-skills, MCP servers, and an LLM integration benchmark explicitly instructing agents to load skills first. missing for 10: independent third-party (non-Paddle) confirmation that an agent successfully consumed llms.txt end-to-end in the wild.

   • [claimed-docs] “We support four official server-side SDKs: Node.js, Python, Go, and PHP.”
   • [probe] “PROBE llms.txt: HTTP 200 at https://developer.paddle.com/llms.txt # Paddle Developer Docs > Official developer documentation for Paddle Bil…”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [claimed-docs] “Paddle publishes an LLM integration benchmark as agent-facing guidance: "Real-world test of how well LLMs build with Paddle. 23 models teste…”
   • [claimed-docs] “Connect AI agents like Claude and Cursor directly to Paddle to work through complex flows in seconds. Build pricing models, investigate fail…”

   proof ↗⚑ flag

2. ai-native userRun the product headlessly / in CI for automation

   weight 2 · round drawn
   Adyen~partialclaimed6/10

   Adyen's core integration is a server-to-server REST API (docs-1, docs-2) that can be called entirely headlessly with no UI dependency, supports idempotency keys for safe automated retries (adyen-docs-4, adyen-docs-25), and provides a dedicated test environment with test credentials explicitly meant for pre-production integration testing (adyen-docs-6, adyen-docs-18). Scoped, multiple API credentials support automated/service-account style access (adyen-docs-26, adyen-supp-api-credential-scoping). However, there is no explicit CI/CD pipeline documentation, no CLI/SDK testing harness, and the MCP server is documented as running locally via npx rather than as a CI-oriented headless service. Missing for 10: explicit CI/CD integration guides, automated test-suite/SDK tooling, and confirmation of non-interactive credential provisioning for pipelines.

   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”
   • [claimed-docs] “Your server makes three Checkout API requests: /paymentMethods, /payments, and /payments/details.”
   • [claimed-docs] “To submit a request for idempotent processing, send a request with the `idempotency-key:` in the header.”
   • [claimed-docs] “The Adyen API supports idempotency, allowing you to retry a request multiple times while only performing the action once.”
   • [claimed-docs] “We recommend testing your full integration, including the client-side and server-side parts, in your test environment before accepting live …”
   • [claimed-docs] “Test the different types of transactions with Adyen that your integration handles using test credentials.”
   • [claimed-docs] “You can also create multiple API credentials to improve security and control access.”
   • [claimed-docs] “Adyen API credentials are scoped by construction: each credential carries "Roles: Permissions that define what the credential is allowed to …”

   proof ↗⚑ flag

   Paddle~partialprobed6/10

   Paddle exposes a full REST API with Bearer-token auth, OpenAPI spec, Postman collection, and server-side SDKs (Node/Python/Go/PHP) that can all be scripted headlessly in CI, plus a documented Sandbox environment for automated testing and webhooks for event-driven automation. However, this evidence is all API/SDK automation rather than a CLI or headless mode for the product's own dashboard/UI workflows, and there's no explicit CI pipeline example, GitHub Actions integration, or independent hands-on confirmation of running Paddle in CI. missing for 10: dedicated CLI tool, documented CI/CD pipeline examples (e.g., GitHub Actions), independent/community verification of headless CI usage.

   • [claimed-docs] “Paddle publishes an OpenAPI 3.1 specification for the Paddle Billing API. Use it to generate client libraries, drive editor autocomplete, or…”
   • [claimed-docs] “Use Bearer authentication when making requests to the Paddle API. Get a key from Paddle > Developer tools > Authentication.”
   • [claimed-docs] “"Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments..…”
   • [claimed-docs] “We support four official server-side SDKs: Node.js, Python, Go, and PHP.”
   • [claimed-docs] “Import the Paddle Postman collection to explore and test the Paddle Billing API interactively.”
   • [claimed-docs] “Paddle uses webhooks to push real-time notifications to your server when things happen in your account.”
   • [probe] “PROBE llms.txt: HTTP 200 at https://developer.paddle.com/llms.txt # Paddle Developer Docs > Official developer documentation for Paddle Bil…”

   proof ↗⚑ flag

3. ai-native userConnect an agent via an official MCP server

   weight 3 · round to Paddle
   Adyen✓fullprobed8/10

   Adyen documents and ships an official MCP server (github.com/Adyen/adyen-mcp) that lets LLM clients call Checkout, payment links, modifications, and Management API tools via natural language, with clear setup (npx install), scoped API-key auth, and role-based permissions. Missing for 10: independent/hands-on third-party verification of the MCP server working in practice (only vendor docs and a live-probe confirming the doc page exists, no community usage reports) and clarity on production-readiness given its 'Alpha' status.

   • [claimed-docs] “Our Model Context Protocol (MCP) server lets you use natural language with your chosen Large Language Model (LLM) client.”
   • [claimed-docs] “This allows developers to use natural language to connect with Adyen's platform more easily and build new payment solutions faster.”
   • [probe] “official MCP server documented at https://docs.adyen.com/development-resources/mcp-server”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “MCP access rides on Adyen's standard scoped credentials: "To authenticate the requests you will make using the MCP, generate an API key. Mak…”

   proof ↗⚑ flag

   Paddle✓fullprobed9/10

   Paddle documents an official, hosted, remote MCP server (sandbox and live endpoints, OAuth/API-key auth) that lets agents like Claude and Cursor perform real account actions—pricing, billing, refunds, subscription changes—plus a separate docs MCP server and agent skills, all corroborated by an independent probe confirming the docs URL. This is a first-party, production MCP integration well beyond a generic API wrapper. Missing for 10: independent hands-on user report of connecting an agent to the live MCP server (only vendor docs and a URL probe, no third-party usage account).

   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “A separate docs MCP server connects "AI agents to current Paddle documentation, the OpenAPI specification, and SDK references" (hosted at pa…”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [claimed-docs] “Connect AI agents like Claude and Cursor directly to Paddle to work through complex flows in seconds.”
   • [claimed-docs] “The Paddle MCP server lets AI agents interact with your Paddle account. Instead of navigating the dashboard or making API calls, you can ask…”
   • [probe] “official MCP server documented at https://developer.paddle.com/sdks/ai/paddle-mcp”

   proof ↗⚑ flag

4. ai-native userUse an official CLI

   weight 2 · round drawn
   Adyen—none0/10

   Evidence covers an MCP server, OpenAPI specs, docs tooling, and agentic-commerce protocols, but there is no mention of an official Adyen CLI for AI-native workflows anywhere in the pack.

   ⚑ flag

   Paddle—none0/10

   Paddle's evidence covers SDKs, MCP servers, agent skills, OpenAPI/Postman specs, and webhooks, but nowhere documents an official Paddle CLI tool; the mentions of 'Claude Code, Codex, and Gemini CLI' refer to third-party AI coding CLIs that Paddle plugins integrate with, not a Paddle-branded CLI.

   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [claimed-docs] “We support four official server-side SDKs: Node.js, Python, Go, and PHP.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://developer.paddle.com/openapi.json, https://developer.paddle.com/swagger.json, https://develo…”

   proof ↗⚑ flag

5. ai-native userDrive the product through a documented public API

   weight 3 · round drawn
   Adyen✓fullprobed9/10

   Adyen ships extensive documented public REST APIs (Checkout, Management, Platforms, Payouts, Agentic Commerce, etc.), an OpenAPI spec repo as 'single source of truth' for client generation, idempotency keys, scoped API credentials, webhooks, and machine-readable docs (llms.txt, .md pages, full docs mirror) explicitly aimed at AI tooling. Probes independently confirm llms.txt and .md doc endpoints return 200. Missing for 10: a live discoverable OpenAPI/swagger endpoint directly under docs.adyen.com (probe found 404s on standard paths) and independent third-party corroboration of API usability beyond vendor docs.

   • [claimed-docs] “The Adyen OpenAPI Specification repository (github.com/Adyen/adyen-openapi) "contains OpenAPI specifications for every public API Adyen offe…”
   • [claimed-docs] “"Adyen makes documentation accessible to developers as well as the software and AI tools they create." Documented agent-docs surface: llms.t…”
   • [claimed-docs] “The Adyen API supports idempotency, allowing you to retry a request multiple times while only performing the action once.”
   • [claimed-docs] “Adyen API credentials are scoped by construction: each credential carries "Roles: Permissions that define what the credential is allowed to …”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process, or for…”
   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.adyen.com/llms.txt # Adyen Docs > Developer and merchant documentation for Adyen payments, Adyen f…”
   • [probe] “PROBE docs-md: HTTP 200 at https://docs.adyen.com/online-payments.md --- title: "Online payments" description: "Accept payments with cards, …”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.adyen.com/openapi.json, https://docs.adyen.com/swagger.json, https://docs.adyen.com/api…”

   proof ↗⚑ flag

   Paddle✓fullprobed9/10

   Paddle exposes a full public REST API with Bearer auth, granular permissions, key rotation, an OpenAPI 3.1 spec, Postman collection, and official SDKs (Node, Python, Go, PHP) — plus a purpose-built MCP server with codemode search/execute tools, agent skills, and a docs MCP, explicitly designed for AI agents to drive the product end-to-end. Missing for 10: independent third-party corroboration of the MCP/API working reliably at scale (evidence is entirely vendor docs), and the probe found the OpenAPI spec not resolvable at common guessed paths.

   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “Machine-readable API surface: "Download the Paddle OpenAPI specification to generate client libraries, power editor tooling, or import into …”
   • [claimed-docs] “Use Bearer authentication when making requests to the Paddle API. Get a key from Paddle > Developer tools > Authentication.”
   • [claimed-docs] “Permissions grant an API key access to specific entities or allow it to take specific actions in Paddle.”
   • [claimed-docs] “Paddle publishes an OpenAPI 3.1 specification for the Paddle Billing API. Use it to generate client libraries, drive editor autocomplete, or…”
   • [claimed-docs] “Bearer-token API keys with documented granular permissions: keys are created in "Paddle > Developer tools > Authentication", scoped per the …”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [probe] “official MCP server documented at https://developer.paddle.com/sdks/ai/paddle-mcp”

   proof ↗⚑ flag

6. ai-native userIssue scoped/least-privilege API credentials for an agent

   weight 2 · round to Adyen
   Adyen✓fullclaimed8/10

   Adyen documents scoped, role-based API credentials (roles define exactly what a credential can do, e.g. limiting to initiating vs capturing payments, per-merchant-account scoping) and explicitly recommends creating multiple credentials for finer-grained access control — and the MCP server docs instruct assigning a dedicated API credential with only the specific roles needed for agent use, plus a --tools flag to limit which MCP tools are exposed. This directly supports least-privilege credential issuance for an AI agent. missing for 10: no independent/hands-on verification of least-privilege enforcement in practice, and no explicit per-agent short-lived/ephemeral credential mechanism beyond static role-scoped API keys.

   • [claimed-docs] “Adyen API credentials are scoped by construction: each credential carries "Roles: Permissions that define what the credential is allowed to …”
   • [claimed-docs] “MCP access rides on Adyen's standard scoped credentials: "To authenticate the requests you will make using the MCP, generate an API key. Mak…”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “You can also create multiple API credentials to improve security and control access.”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle supports API keys with documented granular permissions/scopes (paddle-docs-9, paddle-supp-auth-permissions) and sandbox-only keys (paddle-supp-sandbox) that can be issued for agent use via the MCP server, and Live OAuth MCP connections default to read-only access initially (paddle-supp-mcp-server). However, the MCP server itself explicitly does not gate destructive operations and relies on the user manually creating least-privilege keys rather than the product enforcing scoped credential issuance for agents (paddle-supp-mcp-codemode). missing for 10: no dedicated agent-specific credential/token issuance flow (e.g. per-agent scoped key generation UI or API), no evidence of fine-grained action-level restriction enforced automatically for MCP/agent use, no independent confirmation that permission scoping actually restricts agent behavior in practice.

   • [claimed-docs] “Permissions grant an API key access to specific entities or allow it to take specific actions in Paddle.”
   • [claimed-docs] “Bearer-token API keys with documented granular permissions: keys are created in "Paddle > Developer tools > Authentication", scoped per the …”
   • [claimed-docs] “"Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments..…”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”

   proof ↗⚑ flag

7. ai-native userBuild against official SDKs

   weight 2 · round to Paddle
   Adyen~partialprobed6/10

   Adyen documents official SDKs for web/iOS/Android/React Native/Flutter and a public OpenAPI spec repo as the 'single source of truth' usable to generate client libraries, which supports SDK-based building; however, the OpenAPI/spec endpoint probes returned 404s and there's no independent/hands-on corroboration of SDK quality or completeness for AI-native workflows specifically. missing for 10: independent verification of official SDK repos/quality, working OpenAPI spec endpoint (probe returned 404s), hands-on developer confirmation of SDK-based agentic integration.

   • [claimed-docs] “Support online payments for web, iOS, Android, React Native, or Flutter.”
   • [claimed-docs] “The Adyen OpenAPI Specification repository (github.com/Adyen/adyen-openapi) "contains OpenAPI specifications for every public API Adyen offe…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.adyen.com/openapi.json, https://docs.adyen.com/swagger.json, https://docs.adyen.com/api…”

   proof ↗⚑ flag

   Paddle✓fullprobed9/10

   Paddle publishes official server-side SDKs (Node.js, Python, Go, PHP) with TypeScript definitions, pagination helpers, and webhook signature verification, alongside a maintained OpenAPI 3.1 spec (GitHub repo) and Postman collection for generating clients — all documented as first-party, developer-facing tooling suited to AI-native/agentic building workflows. Missing for 10: independent hands-on developer corroboration of SDK quality/DX beyond Paddle's own docs, and the direct openapi.json endpoint probe 404'd (spec is hosted elsewhere per docs, a minor discrepancy).

   • [claimed-docs] “We support four official server-side SDKs: Node.js, Python, Go, and PHP.”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps.”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps. It ships with TypeScript definitions, iter…”
   • [claimed-docs] “The Paddle Python SDK integrates Paddle Billing with Python applications. It handles pagination automatically, exposes operation classes for…”
   • [claimed-docs] “Machine-readable API surface: "Download the Paddle OpenAPI specification to generate client libraries, power editor tooling, or import into …”
   • [claimed-docs] “Paddle publishes an OpenAPI 3.1 specification for the Paddle Billing API. Use it to generate client libraries, drive editor autocomplete, or…”
   • [claimed-docs] “Postman is the quickest way to explore and test the Paddle Billing API. The Paddle Postman collection includes every operation, so you can t…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://developer.paddle.com/openapi.json, https://developer.paddle.com/swagger.json, https://develo…”

   proof ↗⚑ flag

8. ai-native userSubscribe to events via webhooks

   weight 2 · round drawn
   Adyen✓fullclaimed8/10

   Adyen provides first-party webhook subscriptions to avoid polling, with event logs visible in the Customer Area, and MCP tools to list/get/test webhooks, plus agentic-commerce flows that explicitly send webhook payment outcomes to merchant servers — showing webhooks are usable both directly and via AI-native/agent tooling. missing for 10: no independent/hands-on developer confirmation of webhook subscription setup via API or MCP, and no detail on webhook event types/payload schema management through natural-language/agentic tooling.

   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process, or for…”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process”
   • [claimed-docs] “In your Customer Area, you can find a log of all webhook events sent to your server.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “Google Agentic Commerce integration: "Use this integration for processing Google Pay payments through Adyen, when Google agents (Google AI M…”

   proof ↗⚑ flag

   Paddle✓fullclaimed8/10

   Paddle has robust first-party webhook documentation: real-time push notifications for account events, signature verification via Paddle-Signature header, IP allowlisting guidance, and SDK helpers (Node.js, Python) for verifying webhook signatures. MCP tooling also explicitly supports 'configure webhooks, and simulate events for testing,' making this agent-accessible. Missing for 10: independent/hands-on confirmation of webhook reliability or agent-driven webhook subscription workflows beyond vendor docs.

   • [claimed-docs] “Paddle uses webhooks to push real-time notifications to your server when things happen in your account.”
   • [claimed-docs] “you can use the secret key to generate your own signature for each webhook... if both signatures match then you can be sure that a received …”
   • [claimed-docs] “All webhooks sent by Paddle include a `Paddle-Signature` header. Paddle generates this header using a secret key that only you and Paddle kn…”
   • [claimed-docs] “We recommend adding Paddle webhook IP addresses to your allowlist, and rejecting webhooks that come from other sources.”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps. It ships with TypeScript definitions, iter…”
   • [claimed-docs] “The Paddle Python SDK integrates Paddle Billing with Python applications. It handles pagination automatically, exposes operation classes for…”
   • [claimed-docs] “Generate integration code, configure webhooks, and simulate events for testing.”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”

   proof ↗⚑ flag

Agentic features

1. ai-native userGet AI-generated insights and suggestions from my data inside the product

   weight 2 · round to Adyen
   Adyen~partialclaimed6/10

   Adyen Uplift is documented as an AI layer that analyzes payment data and 'gives you recommendations that are tailored for your business' to optimize conversion, risk, and cost, and risk management includes ML-powered fraud detection using global transaction data. This matches the spirit of AI-generated insights/suggestions from account data, but it's narrowly scoped to payment-funnel optimization and fraud, not a general-purpose 'insights' feature across reporting/dashboards, and there's no independent/hands-on corroboration of how these suggestions surface in the UI. Missing for 10: broader AI-generated insights across reporting/analytics (not just Uplift/fraud), and third-party or hands-on evidence of the insights UX.

   • [claimed-docs] “Adyen Uplift is the platform's AI optimization layer: "Adyen Uplift helps you strike a balance between conversion, risk and cost. You get re…”
   • [claimed-docs] “Adyen Uplift helps you strike a balance between conversion, risk and cost. You get recommendations that are tailored for your business”
   • [claimed-docs] “Create, backtest, and label custom rules with a rule builder interface | Machine learning powered fraud detection using global transaction d…”
   • [claimed-docs] “Machine learning powered fraud detection using global transaction data | | | Create, backtest, and label custom rules with a rule builder in…”
   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”

   proof ↗⚑ flag

   Paddle~partialclaimed4/10

   Paddle's MCP server explicitly supports 'analyze revenue, refunds, and transaction patterns' and 'investigate failed payments' via natural-language queries from a connected AI agent (Claude, Cursor, etc.), which can surface insights from account data. However, this requires an external agent/tool to connect via MCP rather than an in-product, always-on AI insights/suggestions feature (e.g., no evidence of a native dashboard widget proactively surfacing AI-generated recommendations). Missing for 10: evidence of built-in, in-app AI-generated insights/suggestions surfaced automatically to users without requiring a separate agent connection, and independent/hands-on corroboration of the MCP-driven analysis actually producing useful insights in practice.

   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “Process subscription upgrades, downgrades, pauses, and cancellatio”
   • [claimed-docs] “Build pricing models, investigate failed payments, integrate Paddle features, or process customer refunds — all in natural language.”
   • [claimed-docs] “Filter, break down, and investigate the metrics and data available in your Paddle account from the dashboard or the API.”

   proof ↗⚑ flag

2. ai-native userSet up automations that run autonomously in the background

   weight 2 · round to Adyen
   Adyen~partialclaimed5/10

   Adyen supports several background/event-driven automations — webhooks that eliminate manual polling for async status changes, Uplift's AI-driven payment-funnel optimization experiments that can be started/evaluated/stopped autonomously, and risk-management rules/ML fraud detection that run continuously — but these are fixed platform features rather than a general-purpose mechanism for an AI-native user to author custom autonomous background automations. Missing for 10: a user-configurable automation/workflow builder or scheduler that an AI-native user can set up to run arbitrary agentic tasks unattended, and independent evidence of such automations running reliably in production.

   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process, or for…”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process”
   • [claimed-docs] “Adyen Uplift is the platform's AI optimization layer: "Adyen Uplift helps you strike a balance between conversion, risk and cost. You get re…”
   • [claimed-docs] “Create, backtest, and label custom rules with a rule builder interface | Machine learning powered fraud detection using global transaction d…”
   • [claimed-docs] “Machine learning powered fraud detection using global transaction data | | | Create, backtest, and label custom rules with a rule builder in…”

   proof ↗⚑ flag

   Paddle—none0/10

   Paddle's evidence covers webhooks, MCP servers, and agent skills for integration purposes, but none of this describes autonomous background automations (e.g., scheduled/triggered workflows that run without user invocation) — the product is a payments/billing platform with agent-assisted integration tooling, not an automation/workflow engine that runs tasks autonomously in the background. missing for 10: any evidence of scheduled jobs, autonomous triggers, or background automation execution initiated by the user rather than by an explicit API/webhook call.

   ⚑ flag

3. ai-native userDelegate tasks to a built-in AI assistant inside the product

   weight 3 · round to Paddle
   Adyen—none0/10

   Adyen documents an MCP server (for external LLM clients to call Adyen's APIs) and 'Uplift' (automated AI-driven payment optimization), but no evidence of a built-in AI assistant/copilot inside the Adyen product itself that a user can converse with or delegate tasks to interactively.

   • [claimed-docs] “Our Model Context Protocol (MCP) server lets you use natural language with your chosen Large Language Model (LLM) client.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “Adyen Uplift is the platform's AI optimization layer: "Adyen Uplift helps you strike a balance between conversion, risk and cost. You get re…”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle doesn't ship a native, in-dashboard AI assistant chat UI; instead it exposes a hosted MCP server and agent-skills that let external AI agents (Claude, Cursor, etc.) connect to Paddle and perform pricing, billing, and integration tasks in natural language, plus a codemode search/execute interface for chaining API calls. This delivers the 'delegate tasks to AI' outcome but only via a third-party agent client rather than an assistant embedded inside Paddle's own product surface. Missing for 10: a first-party in-product chat/assistant UI (not just an MCP endpoint for external agents), and independent hands-on evidence of the delegation experience.

   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The Paddle MCP server lets AI agents interact with your Paddle account. Instead of navigating the dashboard or making API calls, you can ask…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [claimed-docs] “Build pricing models, investigate failed payments, integrate Paddle features, or process customer refunds — all in natural language.”

   proof ↗⚑ flag

4. ai-native userOperate the product with natural-language commands

   weight 2 · round to Paddle
   Adyen~partialprobed7/10

   Adyen ships an official MCP server (adyen-mcp) that explicitly lets users 'use natural language with your chosen LLM client' to interact with Adyen APIs (create sessions, payment links, refunds, management API), with documented scoped-credential setup — a genuine natural-language operation surface for AI-native users. However, the MCP server is explicitly marked 'Alpha', runs only locally via npx, and only exposes a documented subset of tools rather than full product coverage. Missing for 10: production-grade/GA status, independent hands-on validation of the NL interface actually working reliably, and broader coverage beyond the documented tool subset.

   • [claimed-docs] “Our Model Context Protocol (MCP) server lets you use natural language with your chosen Large Language Model (LLM) client.”
   • [claimed-docs] “This allows developers to use natural language to connect with Adyen's platform more easily and build new payment solutions faster.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “MCP access rides on Adyen's standard scoped credentials: "To authenticate the requests you will make using the MCP, generate an API key. Mak…”
   • [probe] “official MCP server documented at https://docs.adyen.com/development-resources/mcp-server”

   proof ↗⚑ flag

   Paddle✓fullclaimed9/10

   Paddle documents an extensive MCP-based natural-language interface: the hosted Paddle MCP server lets agents 'ask an agent in natural language to handle pricing, billing, support, and integration workflows,' covering pricing changes, refunds, subscription upgrades/downgrades/cancellations, webhook config, and report generation, backed by a codemode search/execute interface, agent skills, and a published LLM benchmark showing measured task success. This is rich, detailed first-party documentation directly on-topic for natural-language operation. Missing for 10: independent/hands-on third-party corroboration that the natural-language commands work reliably in practice (community evidence available is unrelated to AI/MCP usage).

   • [claimed-docs] “Connect AI agents like Claude and Cursor directly to Paddle to work through complex flows in seconds.”
   • [claimed-docs] “Connect AI agents like Claude and Cursor directly to Paddle to work through complex flows in seconds. Build pricing models, investigate fail…”
   • [claimed-docs] “The Paddle MCP server lets AI agents interact with your Paddle account. Instead of navigating the dashboard or making API calls, you can ask…”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [claimed-docs] “Paddle publishes an LLM integration benchmark as agent-facing guidance: "Real-world test of how well LLMs build with Paddle. 23 models teste…”
   • [claimed-docs] “Lets AI agents read data and take actions in your Paddle account.”

   proof ↗⚑ flag

Api quality

1. ai-native userExplore an interactive API reference with runnable examples

   weight 2 · round to Paddle
   Adyen—none0/10

   The evidence describes OpenAPI specs (usable to generate Postman collections or client libraries), markdown-rendered docs, and an MCP server for natural-language API access, but there is no evidence of an actual interactive API reference/console embedded in Adyen's docs where a user can run live example requests. A probe for a standard OpenAPI/Swagger endpoint also returned 404s, and no 'try it' console or embedded runnable code examples are documented.

   • [claimed-docs] “The Adyen OpenAPI Specification repository (github.com/Adyen/adyen-openapi) "contains OpenAPI specifications for every public API Adyen offe…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.adyen.com/openapi.json, https://docs.adyen.com/swagger.json, https://docs.adyen.com/api…”
   • [claimed-docs] “"Adyen makes documentation accessible to developers as well as the software and AI tools they create." Documented agent-docs surface: llms.t…”

   proof ↗⚑ flag

   Paddle~partialprobed6/10

   Paddle documents an OpenAPI 3.1 spec and an importable Postman collection explicitly framed for interactive exploration/testing ('try the API interactively before writing code', 'explore and test the Paddle Billing API interactively'), which supports runnable-example exploration for AI-native/developer users. However, a probe for a directly hosted interactive reference page (Swagger/Redoc-style) on the docs site returned 404s at common paths, suggesting the 'interactive reference' is delegated to external tools (Postman, Scalar, Redocly) rather than a first-party built-in interactive console. Missing for 10: confirmed first-party hosted interactive API console with live runnable examples, and independent hands-on corroboration of the interactive experience.

   • [claimed-docs] “Import the Paddle Postman collection to explore and test the Paddle Billing API interactively.”
   • [claimed-docs] “Paddle publishes an OpenAPI 3.1 specification for the Paddle Billing API. Use it to generate client libraries, drive editor autocomplete, or…”
   • [claimed-docs] “Postman is the quickest way to explore and test the Paddle Billing API. The Paddle Postman collection includes every operation, so you can t…”
   • [claimed-docs] “Machine-readable API surface: "Download the Paddle OpenAPI specification to generate client libraries, power editor tooling, or import into …”
   • [probe] “PROBE openapi: all candidate paths 404 (https://developer.paddle.com/openapi.json, https://developer.paddle.com/swagger.json, https://develo…”

   proof ↗⚑ flag

2. ai-native userDownload a machine-readable API spec (OpenAPI or equivalent)

   weight 2 · round drawn
   Adyen✓fullprobed8/10

   Adyen documents a dedicated OpenAPI specification repository (github.com/Adyen/adyen-openapi) covering 'every public API Adyen offers' as the 'single source of truth,' explicitly positioned for machine consumption (client-library generation, Postman collections, and LLM/agent context). This is reinforced by llms.txt, markdown-renderable docs, and a full docs mirror ZIP for AI tooling. missing for 10: independent/hands-on confirmation that the OpenAPI repo is current and complete, and a direct in-docs discoverable openapi.json endpoint (probe found only 404s for openapi.json-style paths on the docs domain itself).

   • [claimed-docs] “The Adyen OpenAPI Specification repository (github.com/Adyen/adyen-openapi) "contains OpenAPI specifications for every public API Adyen offe…”
   • [claimed-docs] “"Adyen makes documentation accessible to developers as well as the software and AI tools they create." Documented agent-docs surface: llms.t…”
   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.adyen.com/llms.txt # Adyen Docs > Developer and merchant documentation for Adyen payments, Adyen f…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.adyen.com/openapi.json, https://docs.adyen.com/swagger.json, https://docs.adyen.com/api…”

   proof ↗⚑ flag

   Paddle✓fullprobed8/10

   Paddle publishes an OpenAPI 3.1 spec maintained in a public GitHub repo (github.com/PaddleHQ/paddle-openapi) documented for generating client libraries and editor tooling, plus a Postman collection, giving AI-native users a downloadable machine-readable spec. However, a live probe found the spec not hosted at common guessable URLs on developer.paddle.com itself, so discoverability outside docs pages is imperfect. missing for 10: direct probe confirmation of a working download URL, independent third-party corroboration of the spec's completeness/currency.

   • [claimed-docs] “Paddle publishes an OpenAPI 3.1 specification for the Paddle Billing API. Use it to generate client libraries, drive editor autocomplete, or…”
   • [claimed-docs] “Paddle publishes an OpenAPI 3.1 specification for the Paddle Billing API. Use it to generate client libraries, drive editor autocomplete, or…”
   • [claimed-docs] “Machine-readable API surface: "Download the Paddle OpenAPI specification to generate client libraries, power editor tooling, or import into …”
   • [probe] “PROBE openapi: all candidate paths 404 (https://developer.paddle.com/openapi.json, https://developer.paddle.com/swagger.json, https://develo…”

   proof ↗⚑ flag

3. ai-native userTest against a sandbox environment without touching production data

   weight 1 · round to Paddle
   Adyen✓fullclaimed8/10

   Adyen documents a dedicated test environment for the full integration (client- and server-side) using test credentials before going live, and this test environment is explicitly wired into AI-native workflows via the Adyen MCP server which runs with `--env=TEST` (or LIVE with a live prefix), letting an AI agent exercise Checkout, payment links, and modification tools against sandbox data. Missing for 10: no independent/hands-on confirmation that sandbox test data is fully isolated from production, and no detail on seeding/resetting sandbox state for repeatable AI-driven test runs.

   • [claimed-docs] “We recommend testing your full integration, including the client-side and server-side parts, in your test environment before accepting live …”
   • [claimed-docs] “Test the different types of transactions with Adyen that your integration handles using test credentials.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “Our Model Context Protocol (MCP) server lets you use natural language with your chosen Large Language Model (LLM) client.”

   proof ↗⚑ flag

   Paddle✓fullclaimed9/10

   Paddle documents a dedicated Sandbox environment with fully separate datasets, credentials, dashboards and even a distinct sandbox MCP endpoint/API key prefix, explicitly for building and testing integrations without touching real data or payments — directly serving an AI agent's need to safely test flows. Agent skills and MCP server docs also explicitly reference sandbox testing workflows. Missing for 10: independent/hands-on confirmation that AI agents actually operate correctly against sandbox-mcp.paddle.com in practice (only vendor docs, no third-party corroboration).

   • [claimed-docs] “Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments.”
   • [claimed-docs] “"Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments..…”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”

   proof ↗⚑ flag

4. ai-native userRely on versioned APIs with a documented deprecation policy

   weight 2 · round drawn
   Adyen—none0/10

   The evidence pack covers API idempotency, webhooks, credentials, OpenAPI specs, and agentic commerce integrations, but contains no documentation of API versioning scheme or a deprecation policy for older API versions. Versioning/deprecation policy is a standard and fair axis for a payments API platform, so absence of evidence yields 'none'.

   ⚑ flag

   Paddle—none0/10

   The evidence pack covers OpenAPI specs, SDKs, authentication, and a changelog, but nowhere documents an explicit API versioning scheme or a deprecation policy for breaking changes. Missing for 10: documented API version headers/scheme, explicit deprecation/sunset policy, and migration timelines for breaking changes.

   ⚑ flag

1. ai-native userPerform bulk operations across many items at once

   weight 2 · round to Adyen
   Adyen~partialclaimed6/10

   Adyen documents concrete bulk operations in its agentic Product Feed API — batch ingest of up to 200 items per call and batch delete of up to 200 products per call — which is genuine 'many items at once' functionality for AI-platform inventory sync. However, evidence of bulk operations elsewhere in the platform (e.g., bulk payouts, bulk refunds/modifications, bulk reporting exports, or MCP tools operating on many resources at once) is absent; the MCP tool list is single-item oriented (create one session, one payment link, one refund). Missing for 10: bulk support evidenced across other core domains (payments, payouts, platforms/onboarding) and bulk-capable MCP tools, not just the product feed.

   • [claimed-docs] “You can upload batches of information about your inventory to Adyen, and we will synchronize it with AI platforms.”
   • [claimed-docs] “The Adyen Product Feed API is "a centralized product knowledge hub that synchronizes your inventory with AI platforms" via asynchronous batc…”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”

   proof ↗⚑ flag

   Paddle~partialclaimed5/10

   The MCP 'execute' tool lets an agent chain multiple Paddle API calls in a single invocation, and SDKs offer iterator-based pagination for reading many records, which supports scripted bulk-style workflows for AI agents. However, there is no documented bulk/batch API endpoint (e.g., bulk update prices, bulk refunds) or evidence of handling large-scale operations reliably. Missing for 10: explicit bulk/batch endpoints, evidence of rate-limit handling for large batch jobs, and hands-on proof of bulk operations succeeding at scale.

   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps. It ships with TypeScript definitions, iter…”
   • [claimed-docs] “The Paddle Python SDK integrates Paddle Billing with Python applications. It handles pagination automatically, exposes operation classes for…”
   • [claimed-docs] “Permissions grant an API key access to specific entities or allow it to take specific actions in Paddle.”

   proof ↗⚑ flag

2. ai-native userDefine rules that trigger actions automatically on events

   weight 3 · round to Adyen
   Adyen~partialclaimed5/10

   Adyen documents a genuine rules mechanism — the risk management rule builder lets you 'create, backtest, and label custom rules' that automatically trigger block/allow/review actions on each transaction event (adyen-docs-20, adyen-docs-28, adyen-docs-48), and webhooks notify on async status-change events (adyen-docs-3/17). This satisfies the core 'rule triggers action on event' pattern, but only within the fraud/risk-decisioning domain, not as a general-purpose automation/workflow engine, and there's no evidence of AI-native rule authoring (e.g., via the MCP server or natural language) for these rules. Missing for 10: general cross-domain event-action automation beyond fraud rules, AI/LLM-driven rule creation or triggering, and independent evidence of real-world use of the rule builder.

   • [claimed-docs] “Create, backtest, and label custom rules with a rule builder interface | Machine learning powered fraud detection using global transaction d…”
   • [claimed-docs] “Machine learning powered fraud detection using global transaction data | | | Create, backtest, and label custom rules with a rule builder in…”
   • [claimed-docs] “Each payment goes through a risk evaluation. The risk evaluation reaches a decision that results in an action to block, allow, or review the…”
   • [claimed-docs] “You can detect, monitor, and mitigate fraud using Protect, Adyen's risk management system.”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process, or for…”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process”

   proof ↗⚑ flag

   Paddle~partialclaimed4/10

   Paddle exposes event-driven webhooks that fire on billing events (paddle-docs-11, paddle-docs-45) and lets users configure some built-in automated behaviors like payment recovery/dunning rules and tactical retries (paddle-supp-dunning, paddle-docs-18), which count as rudimentary 'if event X happens, do Y' automation. However, this is limited to a handful of pre-built scenarios (dunning, webhook delivery) rather than a general-purpose rules engine where an AI-native user can define arbitrary custom triggers/actions; actually triggering an action still requires the developer to write server-side code to consume the webhook. Missing for 10: a native no-code/rules-builder UI for arbitrary event-to-action automation, evidence of user-defined conditional logic beyond dunning settings, and any AI-agent-specific example of autonomously creating such rules.

   • [claimed-docs] “Paddle uses webhooks to push real-time notifications to your server when things happen in your account.”
   • [claimed-docs] “We recommend adding Paddle webhook IP addresses to your allowlist, and rejecting webhooks that come from other sources.”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”
   • [claimed-docs] “You can configure how Payment Recovery notifies customers about failed payments, turn on Tactical Retries, and choose what happens to past d…”

   proof ↗⚑ flag

Ai billing ops

1. ai-native userHave an agent monitor failed payments via API and webhooks and drive retries and dunning outreach

   weight 3 · round to Paddle
   Adyen~partialclaimed5/10

   Adyen provides the raw building blocks an agent could use—webhooks for status changes (adyen-docs-3/17/45), refusal-reason detail on failed payments (adyen-docs-7), idempotent retries (adyen-docs-4/25), and an MCP server exposing webhook list/get/test and payment tools (adyen-supp-mcp-tools)—but there is no documented dunning/subscription-retry feature, automated outreach mechanism, or agent workflow specifically for failed-payment recovery. Missing for 10: native dunning/retry orchestration, recurring-billing failed-payment workflow, and evidence of an agent actually driving retries/outreach rather than just monitoring raw webhook/API data.

   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process, or for…”
   • [claimed-docs] “When the payment or modification request is unsuccessful, the `resultCode`, `refusalReason`, and `refusalReasonCode` describe the issue.”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process”
   • [claimed-docs] “The Adyen API supports idempotency, allowing you to retry a request multiple times while only performing the action once.”
   • [claimed-docs] “In your Customer Area, you can find a log of all webhook events sent to your server.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle documents webhooks for real-time payment-failure notifications, an API/MCP server that lets agents 'Investigate failed payments and resolve billing issues,' and built-in dunning (Tactical Retries, notification config) — covering the monitoring and resolution loop an AI agent could drive via API/MCP. However, the dunning/retry mechanism itself is a built-in automated system (configured, not actively 'driven' step-by-step by an agent), and there's no explicit example of an agent initiating a retry or outreach action end-to-end. Missing for 10: a concrete agent-driven retry/outreach workflow example, and clarity on whether MCP actions can actually trigger/override the automatic retry schedule rather than just configure it.

   • [claimed-docs] “Paddle uses webhooks to push real-time notifications to your server when things happen in your account.”
   • [claimed-docs] “You can configure how Payment Recovery notifies customers about failed payments, turn on Tactical Retries, and choose what happens to past d…”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “you can use the secret key to generate your own signature for each webhook... if both signatures match then you can be sure that a received …”

   proof ↗⚑ flag

2. ai-native userHave an agent provision a new product, price, and payment link via the API in a single session

   weight 3 · round to Paddle
   Adyen—none0/10

   Adyen's API and MCP server cover payments, payment links, checkout sessions, and payouts, but there is no evidence of any 'product' or 'price' resource concept (billing/catalog objects like Stripe's Products/Prices) that an agent could provision alongside a payment link — Adyen is a payments processor, not a billing/invoicing product catalog system. The only 'product' references are to inventory feeds for agentic commerce (syncing existing catalog to AI platforms), not creating new products/prices via API. missing for 10: evidence of a Products/Prices API or equivalent billing object model, evidence of a single-session workflow that creates a product+price+payment link together, any documentation of provisioning billing entities (not just payment links) via API or MCP.

   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “Offer payment links.”
   • [claimed-docs] “The Adyen Product Feed API is "a centralized product knowledge hub that synchronizes your inventory with AI platforms" via asynchronous batc…”

   proof ↗⚑ flag

   Paddle✓fullclaimed8/10

   Paddle's hosted MCP server gives agents a codemode interface (search + execute) that can chain arbitrary Paddle API calls — including creating products, prices, and payment/checkout links — all within one authenticated session, backed by API key/OAuth auth and full OpenAPI coverage (paddle-supp-mcp-server, paddle-supp-mcp-codemode, paddle-docs-15, paddle-docs-46). This directly supports an agent provisioning product+price+payment link end-to-end via API in one session. missing for 10: no independent/hands-on report of an agent actually completing this exact three-step flow (product→price→link) in one session, and no explicit example transcript showing payment-link creation specifically via the MCP execute tool.

   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “The MCP server "uses a codemode interface": three tools cover the full API — `search` ("Looks up Paddle API methods by name or description. …”
   • [claimed-docs] “you may also create your own transactions using the API or Paddle dashboard.”
   • [claimed-docs] “Paddle automatically creates transactions for subscription lifecycle events and when checkouts are opened, but you can create your own trans…”
   • [claimed-docs] “Use Bearer authentication when making requests to the Paddle API. Get a key from Paddle > Developer tools > Authentication.”
   • [claimed-docs] “"Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments..…”

   proof ↗⚑ flag

Invoicing

1. finance-ops userSend hosted invoices with online payment and automatic reminders

   weight 2 · round to Paddle
   Adyen—none0/10

   Adyen's evidence covers payment links, checkout, platforms, payouts, and agentic commerce, but there is no mention of a hosted invoicing product, invoice generation, or automatic payment reminders anywhere in the pack — only generic 'payment links' (adyen-docs-24) which is a different capability.

   • [claimed-docs] “Offer payment links.”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle documents sales-assisted invoicing where a team can 'draft and send invoices for subscriptions, collecting payment manually rather than by charging a card on file' (paddle-docs-17/36), and Paddle automatically creates a subscription once an invoice is issued (paddle-docs-69), implying an online payment path. However, the 'automatic reminders' portion of the story is not clearly evidenced for invoices specifically — the documented dunning/Payment Recovery and Tactical Retries feature (paddle-docs-18, paddle-supp-dunning) applies to failed automatic subscription charges, not to reminders for outstanding hosted invoices. Missing for 10: explicit documentation of automated invoice reminder emails/schedules, and independent/hands-on confirmation that invoice payment links and reminders work as described.

   • [claimed-docs] “Your sales team can draft and send invoices for subscriptions, collecting payment manually rather than by charging a card on file.”
   • [claimed-docs] “Invoices let you offer sales-assisted billing. Your sales team can draft and send invoices for subscriptions, collecting payment manually ra…”
   • [claimed-docs] “Paddle automatically creates a subscription for you once an invoice is issued.”
   • [claimed-docs] “You can configure how Payment Recovery notifies customers about failed payments, turn on Tactical Retries, and choose what happens to past d…”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”

   proof ↗⚑ flag

Subscriptions

1. finance-ops userAutomatically retry failed subscription payments with smart dunning and recovery emails

   weight 2 · round to Paddle
   Adyen—none0/10

   The evidence pack covers payments, payouts, webhooks, idempotency, risk management, platforms/marketplace payouts, and agentic commerce, but contains no mention of subscription billing, dunning management, automatic retry logic for failed recurring payments, or recovery emails. This is a plausible axis for a payments platform, but nothing in the docs or community evidence shows Adyen offering smart dunning/recovery workflows.

   ⚑ flag

   Paddle✓fullclaimed8/10

   Paddle's docs explicitly describe built-in dunning: failed payments are automatically retried up to seven times over 30 days, and Paddle Retain's Payment Recovery adds configurable Tactical Retries, optimized recovery email notifications, and in-app retry forms, plus control over past-due subscription outcomes. missing for 10: independent/hands-on evidence of recovery-email effectiveness or customer testimonials, and the pack notes Retain can't be tested in sandbox, limiting pre-production verification.

   • [claimed-docs] “You can configure how Payment Recovery notifies customers about failed payments, turn on Tactical Retries, and choose what happens to past d…”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”

   proof ↗⚑ flag

2. developerSet up recurring subscription billing with plans, trials, upgrades, and proration

   weight 3 · round to Paddle
   Adyen—none0/10

   Adyen's evidence pack covers payments, payouts, platforms, risk, reporting, and agentic commerce, but there is no mention of subscription/recurring billing constructs like plans, trial periods, upgrades/downgrades, or proration—these are the domain of a billing/subscription-management layer (e.g. Adyen has separate tooling or partners for this) not documented here.

   ⚑ flag

   Paddle✓fullclaimed7/10

   Paddle's docs comprehensively cover setting up subscription checkouts, plans/prices, pausing subscriptions, and upgrading/downgrading (including changing base plan and billing frequency) via replacing products/prices, plus sales-assisted invoicing and dunning for failed payments — a solid recurring-billing foundation. missing for 10: explicit documentation of free-trial period setup and how proration is calculated/applied on plan changes (evidence only implies plan-change mechanics without naming proration).

   • [claimed-docs] “Add a checkout to your app or website using Paddle.js to let customers buy one-time products and sign up for a subscriptions.”
   • [claimed-docs] “Upgrade a subscription Add or remove recurring addons, move from month to annual, or switch plans by”
   • [claimed-docs] “Pause subscriptions to stop billing for them temporarily. Paddle doesn't bill for paused subscriptions until they're resumed.”
   • [claimed-docs] “Customers want to change their base plan, for example going from Pro to Enterprise... Customers want to change their billing frequency, for …”
   • [claimed-docs] “Depending on your billing model, you might let customers upgrade or downgrade their plan. This could mean: Customers want to change their ba…”
   • [claimed-docs] “Upgrade or downgrade subscriptions by replacing items on a subscription.”
   • [claimed-docs] “You can configure how Payment Recovery notifies customers about failed payments, turn on Tactical Retries, and choose what happens to past d…”
   • [claimed-docs] “Your sales team can draft and send invoices for subscriptions, collecting payment manually rather than by charging a card on file.”

   proof ↗⚑ flag

Card acceptance

1. founderAccept an online card payment within a day of signing up

   weight 3 · round to Paddle
   Adyen!disputedcontradicted4/10

   Adyen's docs describe a fast technical integration (single /sessions call, test-then-go-live flow, no setup fees) that could suggest same-day acceptance, but independent community reports concretely contradict a founder's ability to just sign up and start accepting payments quickly: Adyen reportedly refuses small clients ('under the million'), requires large reserves for subscription businesses, and one user describes a multi-hundred-hour integration effort before abandoning it. These are hands-on, specific counter-examples, not generic skepticism, so the 'within a day' claim is disputed rather than simply partial. Missing for 10: any evidence of Adyen actually onboarding a small/new founder-level merchant same-day, or a rebuttal to the reserve/minimum-volume gating reports.

   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”
   • [claimed-docs] “We recommend testing your full integration, including the client-side and server-side parts, in your test environment before accepting live …”
   • [claimed-docs] “Pay per transaction, with no setup or monthly fees.”
   • [community] “Adyen refuses small clients, under the million :/”
   • [community] “One of the bigger drawbacks of Adyen vs Stripe... they require a crazy reserve (in the millions) if your model is subscription based, to be …”
   • [community] “Longtime Stripe user did about a year-long stint with Adyen. Stripe beats Adyen on ease of integration, ease of using the back office, charg…”
   • [community] “Adyen's pricing model looks much more attractive on paper, but the integration complexity and minimum processing volumes usually keep early-…”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle's docs show clear fast paths to a live checkout: overlay checkout requiring 'minimal frontend coding,' Paddle.js embeds, a sandbox for testing without real payments, and ready-made Next.js/Supabase starter kits that bundle checkout + billing in a day-scale setup. However, nothing in the evidence pack addresses the merchant-of-record account verification/KYC timeline that typically gates going live with real card payments, so the 'within a day' claim is not directly substantiated. Missing for 10: explicit onboarding/verification turnaround time, first-party or independent confirmation of same-day live acceptance.

   • [claimed-docs] “Overlay checkout lets you present customers with an overlay that handles all parts of the checkout process — minimal frontend coding require…”
   • [claimed-docs] “Inline checkout lets you embed a checkout and display information about items and totals in your own interface, creating checkout experience…”
   • [claimed-docs] “"Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments..…”
   • [claimed-docs] “Production-ready Next.js SaaS starter kit with Paddle Billing, Supabase auth, and Vercel deployment. Includes a localized pricing page, inte…”
   • [claimed-docs] “By the end, you'll have a working three-tier subscription app integrated with Supabase for auth and Paddle webhooks for syncing customer dat…”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps.”

   proof ↗⚑ flag

Checkout surfaces

1. developerEmbed customizable payment components directly in my own site or app instead of redirecting to a hosted page

   weight 3 · round drawn
   Adyen✓fullcommunity8/10

   Adyen's Web/mobile Drop-in and Components let developers embed sessions-based payment UI directly in their own site/app (web, iOS, Android, React Native, Flutter) with customizable styling and localization, rather than redirecting to a hosted page. This is a core, well-documented capability with sample flows (create session, mount Components) and cross-platform SDK support. Missing for 10: independent hands-on confirmation of embedding ease (community feedback is mixed/critical on integration complexity), and no detailed code-level customization examples in the pack.

   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”
   • [claimed-docs] “Support online payments for web, iOS, Android, React Native, or Flutter.”
   • [claimed-docs] “Customize the styling of your checkout.”
   • [claimed-docs] “Customize the styling of your checkout. Present your payment form in the local language of your shoppers with our localization support.”
   • [claimed-docs] “With the client key, you have: A single key for all your allowed origins in an environment.”
   • [community] “I implemented Adyen for a company I was consulting for. I liked it. Their integration was not as terrible as many here make it seem. It allo…”

   proof ↗⚑ flag

   Paddle✓fullclaimed8/10

   Paddle documents both Inline checkout (embed checkout and totals fully within your own UI/interface) and Overlay checkout, explicitly positioned as alternatives to a hosted redirect page, backed by Paddle.js and localized pricing components. missing for 10: independent/hands-on developer corroboration of embedding customization depth, and no visual/styling API detail beyond docs claims.

   • [claimed-docs] “Inline checkout lets you embed a checkout and display information about items and totals in your own interface, creating checkout experience…”
   • [claimed-docs] “Overlay checkout lets you present customers with an overlay that handles all parts of the checkout process — minimal frontend coding require…”
   • [claimed-docs] “Add a checkout to your app or website using Paddle.js to let customers buy one-time products and sign up for a subscriptions.”
   • [claimed-docs] “Build pricing pages that show customers prices localized for their country and currency — including estimated taxes and discounts.”

   proof ↗⚑ flag

2. developerRedirect customers to a vendor-hosted, conversion-optimized checkout page without building my own payment UI

   weight 3 · round to Paddle
   Adyen~partialclaimed6/10

   Adyen explicitly offers hosted 'payment links' that developers can generate and redirect customers to without building payment UI (adyen-docs-24), and the MCP tool surface confirms this as a first-class API capability (create/get/expire payment links) (adyen-supp-mcp-tools). However, Adyen's primary integration path emphasized in the docs is the Sessions/Components client-side flow which requires building or embedding UI (adyen-docs-1), and there is no detailed documentation on conversion-optimization specifics of the hosted page itself. Missing for 10: dedicated documentation describing the hosted checkout page's design/conversion features, independent evidence of its effectiveness, and clarity on it being the recommended 'no-code UI' path rather than a secondary option.

   • [claimed-docs] “Offer payment links.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”

   proof ↗⚑ flag

   Paddle✓fullclaimed9/10

   Paddle explicitly documents overlay checkout ('present customers with an overlay that handles all parts of the checkout process — minimal frontend coding required') and Paddle.js checkout integration, which is vendor-hosted and conversion-optimized without building a custom payment UI. This is corroborated by supporting docs on inline vs overlay checkout options and sandbox testing. Missing for 10: independent hands-on developer reports validating conversion optimization or checkout UX quality beyond vendor docs.

   • [claimed-docs] “Add a checkout to your app or website using Paddle.js to let customers buy one-time products and sign up for a subscriptions.”
   • [claimed-docs] “Overlay checkout lets you present customers with an overlay that handles all parts of the checkout process — minimal frontend coding require…”
   • [claimed-docs] “Inline checkout lets you embed a checkout and display information about items and totals in your own interface, creating checkout experience…”
   • [claimed-docs] “Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments.”

   proof ↗⚑ flag

No code selling

1. founderCreate a shareable payment link from the dashboard with no code and start selling immediately

   weight 2 · round to Adyen
   Adyen!disputedcontradicted4/10

   Adyen's docs confirm a payment-links feature exists ("Offer payment links" - adyen-docs-24) and a Customer Area dashboard for managing transactions (adyen-docs-16), but there is no evidence of a self-serve, no-code flow a founder can use to instantly start selling. Community reports concretely contradict the 'start immediately' premise for founders: Adyen is described as refusing small clients ('Adyen refuses small clients, under the million' - adyen-comm-1) and requiring underwriting/minimum volumes that keep early-stage startups away ('integration complexity and minimum processing volumes usually keep early-stage startups on Stripe' - adyen-comm-11), unlike a turnkey signup-and-sell experience. missing for 10: documented self-serve onboarding flow for small merchants, step-by-step dashboard payment-link creation, evidence that a founder can go live without a sales/underwriting process.

   • [claimed-docs] “Offer payment links.”
   • [claimed-docs] “Manage your account and transactions from our Customer Area, which shows an overview of all your payment channels.”
   • [community] “Adyen refuses small clients, under the million :/”
   • [community] “Adyen's pricing model looks much more attractive on paper, but the integration complexity and minimum processing volumes usually keep early-…”

   proof ↗⚑ flag

   Paddle—none0/10

   Paddle's evidence centers on developer-integrated checkout (Paddle.js overlay/inline checkout embedded in a website or app) and dashboard-created transactions for internal use, but there is no evidence of a standalone, no-code 'shareable payment link' a founder could generate and immediately distribute (e.g., via email or social media) without any integration work. The closest hints — overlay checkout needing 'minimal frontend coding' and dashboard-created transactions — still imply some setup/integration rather than a pure link-generation flow.

   • [claimed-docs] “Inline checkout lets you embed a checkout and display information about items and totals in your own interface, creating checkout experience…”
   • [claimed-docs] “Overlay checkout lets you present customers with an overlay that handles all parts of the checkout process — minimal frontend coding require…”
   • [claimed-docs] “you may also create your own transactions using the API or Paddle dashboard.”
   • [claimed-docs] “Paddle automatically creates transactions for subscription lifecycle events and when checkouts are opened, but you can create your own trans…”

   proof ↗⚑ flag

Payment methods

1. founderOffer digital wallets and local payment methods (Apple Pay, Google Pay, bank redirects, BNPL) from the same integration

   weight 2 · round to Adyen
   Adyen✓fullcommunity8/10

   Adyen docs explicitly promise 'one integration' covering all preferred payment methods (wallets, local methods, BNPL) via a single Checkout API/sessions flow, with community corroboration that it supports native local methods like iDEAL and Sofort. Missing for 10: an explicit itemized list of Apple Pay/Google Pay/BNPL support and independent hands-on confirmation of wallet/BNPL specifically (only general local-method corroboration exists), plus some community friction reports on integration complexity.

   • [claimed-docs] “Offer your customers all their preferred payment methods with a single integration.”
   • [claimed-docs] “One integration Offer your customers all their preferred payment methods with a single integration.”
   • [claimed-docs] “Support online payments for web, iOS, Android, React Native, or Flutter.”
   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”
   • [claimed-docs] “Your server makes three Checkout API requests: /paymentMethods, /payments, and /payments/details.”
   • [community] “I implemented Adyen for a company I was consulting for. I liked it. Their integration was not as terrible as many here make it seem. It allo…”

   proof ↗⚑ flag

   Paddle~partialclaimed3/10

   Evidence confirms Paddle's broad multi-currency and 200+-country reach, and a specific local payment method (Pix recurring mandate for Brazil) was added, showing the platform does add local payment methods over time. However, there is no explicit documentation of Apple Pay, Google Pay, bank redirects, or BNPL options being available through the same checkout integration. missing for 10: explicit mention/docs of Apple Pay, Google Pay, bank redirect methods, and BNPL support in checkout.

   • [claimed-docs] “Turn on Pix to accept subscription payments from customers in Brazil with a recurring Pix Automático mandate that customers set up at checko…”
   • [claimed-docs] “Paddle has built-in support for multi-currency for both payments and payouts, letting customers purchase in their local currency and letting…”
   • [claimed-docs] “you can unlock new revenue by selling in over 200 countries and territories across the world — no additional setup needed.”
   • [claimed-docs] “With Paddle, you can unlock new revenue by selling in over 200 countries and territories across the world — no additional setup needed.”

   proof ↗⚑ flag

Pci compliance

1. developerKeep my systems out of most of PCI DSS scope because card data is tokenized and never touches my servers

   weight 2 · round to Paddle
   Adyen~partialclaimed5/10

   Adyen's Checkout API/sessions flow shows card data handled client-side and passed via Adyen's hosted fields/client key rather than touching the merchant's server (adyen-docs-1, adyen-docs-44), which implies the tokenization pattern developers rely on for reduced PCI scope. However, none of the evidence explicitly discusses PCI DSS, SAQ levels, or scope-reduction guidance, so the compliance benefit is only inferable from the technical integration pattern, not documented directly. missing for 10: explicit PCI DSS scope-reduction documentation, SAQ-A eligibility guidance, independent compliance certification evidence.

   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”
   • [claimed-docs] “With the client key, you have: A single key for all your allowed origins in an environment.”
   • [claimed-docs] “Support online payments for web, iOS, Android, React Native, or Flutter.”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle's checkout products (overlay and inline via Paddle.js) handle card capture directly, and as merchant of record Paddle also handles tax/compliance — this strongly implies card data never touches the developer's servers, which is the basis for reduced PCI scope. However, the evidence pack never explicitly states PCI DSS compliance or scope reduction, or names a SAQ level, so the compliance-specific claim is inferred rather than documented. missing for 10: explicit PCI DSS compliance/certification statement, explicit scope-reduction language, and independent/third-party confirmation of PCI posture.

   • [claimed-docs] “Inline checkout lets you embed a checkout and display information about items and totals in your own interface, creating checkout experience…”
   • [claimed-docs] “Overlay checkout lets you present customers with an overlay that handles all parts of the checkout process — minimal frontend coding require…”
   • [claimed-docs] “Add a checkout to your app or website using Paddle.js to let customers buy one-time products and sign up for a subscriptions.”
   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”
   • [claimed-docs] “We take care of payments, tax, subscriptions, and metrics with one unified API that does it all.”

   proof ↗⚑ flag

Integration reliability

1. developerSafely retry any API request using idempotency keys so network failures never create duplicate charges

   weight 2 · round to Adyen
   Adyen✓fullclaimed8/10

   Adyen explicitly documents idempotency-key headers for retrying API requests without duplicate side-effects (adyen-docs-4, adyen-docs-25), directly matching the story. Missing for 10: independent/hands-on developer confirmation of idempotency behavior in production and detail on key expiration/edge cases.

   • [claimed-docs] “To submit a request for idempotent processing, send a request with the `idempotency-key:` in the header.”
   • [claimed-docs] “The Adyen API supports idempotency, allowing you to retry a request multiple times while only performing the action once.”

   proof ↗⚑ flag

   Paddle—none0/10

   No evidence in the pack mentions idempotency keys or any mechanism for safely retrying API requests without duplicate charges; the API auth/rate-limit docs cover bearer tokens, permissions, and key rotation but nothing about idempotency handling.

   ⚑ flag

2. developerRely on signed webhook deliveries with automatic retries, event logs, and replay to keep my system consistent with payment state

   weight 3 · round to Paddle
   Adyen~partialclaimed4/10

   Adyen documents webhook event logs in the Customer Area (adyen-docs-45) and general webhook mechanics to avoid polling (adyen-docs-3/17), plus an MCP tool to 'test' webhooks (adyen-supp-mcp-tools) that hints at replay tooling. However, the evidence pack never explicitly documents HMAC/signature verification for webhook payloads, an automatic retry policy for failed webhook deliveries, or a dedicated replay mechanism. Missing for 10: explicit signed-webhook (HMAC) verification docs, documented automatic retry/backoff schedule for webhook delivery, and a concrete replay/redeliver feature beyond the vague MCP 'test' tool.

   • [claimed-docs] “In your Customer Area, you can find a log of all webhook events sent to your server.”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process, or for…”
   • [claimed-docs] “With webhooks, you can avoid having to continuously poll an API endpoint, waiting for a change in status for an asynchronous process”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “To submit a request for idempotent processing, send a request with the `idempotency-key:` in the header.”
   • [claimed-docs] “The Adyen API supports idempotency, allowing you to retry a request multiple times while only performing the action once.”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle documents signed webhooks (Paddle-Signature header + secret-key verification), IP allowlisting guidance, and automatic retries for failed payments/dunning (up to 7 retries over 30 days, Tactical Retries via Retain), plus SDK helpers for signature verification. However, there is no documented event log/dashboard for webhook delivery history, no explicit replay/resend mechanism for missed webhook events, and no mention of delivery guarantees or dead-letter handling. missing for 10: webhook event log/delivery history UI, explicit event replay/resend capability, delivery guarantee/idempotency documentation, independent hands-on confirmation of retry behavior.

   • [claimed-docs] “Paddle uses webhooks to push real-time notifications to your server when things happen in your account.”
   • [claimed-docs] “you can use the secret key to generate your own signature for each webhook... if both signatures match then you can be sure that a received …”
   • [claimed-docs] “All webhooks sent by Paddle include a `Paddle-Signature` header. Paddle generates this header using a secret key that only you and Paddle kn…”
   • [claimed-docs] “We recommend adding Paddle webhook IP addresses to your allowlist, and rejecting webhooks that come from other sources.”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps. It ships with TypeScript definitions, iter…”
   • [claimed-docs] “The Paddle Python SDK integrates Paddle Billing with Python applications. It handles pagination automatically, exposes operation classes for…”

   proof ↗⚑ flag

Onboarding dx

1. developerIntegrate a working payment flow from the docs alone in under a day, with copy-paste examples in my language

   weight 2 · round to Paddle
   Adyen!disputedcontradicted4/10

   Adyen's docs describe a clear-looking integration path (single /sessions call or paymentMethods/payments/payments/details flow, testing, webhooks) and multi-platform SDK support (web, iOS, Android, React Native, Flutter), but no evidence pack item actually shows copy-paste code samples per language, and independent hands-on reports directly contradict the 'under a day' claim: one integrator says 'after hundreds of hours invested, we ditched Adyen' and another states 'Stripe beats Adyen on ease of integration... hands down,' while a third calls it 'much more bare metal' than turnkey competitors. A counter-report says the integration 'was not as terrible as many here make it seem,' so the picture is mixed rather than uniformly negative. missing for 10: actual copy-paste language-specific code samples in the evidence, a documented time-to-first-payment benchmark, and independent confirmation that integration is fast rather than the reported multi-hundred-hour effort.

   • [claimed-docs] “Your server makes a single Checkout API request to the /sessions endpoint. Adyen sends payment data from our server to your client-side appl…”
   • [claimed-docs] “Your server makes three Checkout API requests: /paymentMethods, /payments, and /payments/details.”
   • [claimed-docs] “We recommend testing your full integration, including the client-side and server-side parts, in your test environment before accepting live …”
   • [claimed-docs] “Support online payments for web, iOS, Android, React Native, or Flutter.”
   • [community] “Longtime Stripe user did about a year-long stint with Adyen. Stripe beats Adyen on ease of integration, ease of using the back office, charg…”
   • [community] “I implemented Adyen for a company I was consulting for. I liked it. Their integration was not as terrible as many here make it seem. It allo…”
   • [community] “Adyen is much more 'bare metal.' Think more like a modern Authorize.net. Nobody comes close to Stripe's turnkey developer-friendliness.”
   • [community] “Adyen's pricing model looks much more attractive on paper, but the integration complexity and minimum processing volumes usually keep early-…”

   proof ↗⚑ flag

   Paddle✓fullclaimed8/10

   Paddle's docs include multiple checkout integration paths (inline/overlay), official SDKs for Node.js, Python, Go and PHP with copy-paste style code and TypeScript support, a sandbox for testing without real payments, and a production-ready Next.js SaaS starter kit that yields 'a working three-tier subscription app' with Paddle checkout and webhooks — directly matching the 'working payment flow in under a day' story across languages. missing for 10: independent/hands-on third-party confirmation of the 'under a day' timeframe and broader language coverage beyond the four official SDKs.

   • [claimed-docs] “Inline checkout lets you embed a checkout and display information about items and totals in your own interface, creating checkout experience…”
   • [claimed-docs] “Overlay checkout lets you present customers with an overlay that handles all parts of the checkout process — minimal frontend coding require…”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps.”
   • [claimed-docs] “We support four official server-side SDKs: Node.js, Python, Go, and PHP.”
   • [claimed-docs] “The Paddle Node.js SDK integrates Paddle Billing with server-side JavaScript and TypeScript apps. It ships with TypeScript definitions, iter…”
   • [claimed-docs] “The Paddle Python SDK integrates Paddle Billing with Python applications. It handles pagination automatically, exposes operation classes for…”
   • [claimed-docs] “Production-ready Next.js SaaS starter kit with Paddle Billing, Supabase auth, and Vercel deployment. Includes a localized pricing page, inte…”
   • [claimed-docs] “By the end, you'll have a working three-tier subscription app integrated with Supabase for auth and Paddle webhooks for syncing customer dat…”
   • [claimed-docs] “Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments.”

   proof ↗⚑ flag

Sandbox fidelity

1. developerBuild and test the entire payment flow in a sandbox or test mode that mirrors production, with test cards for every scenario including declines and disputes

   weight 3 · round drawn
   Adyen~partialcommunity6/10

   Adyen documents a dedicated test environment with test credentials for exercising different transaction types (adyen-docs-18), recommends testing the full client+server integration before going live (adyen-docs-6), and documents refusal reasons/result codes for simulating declines (adyen-docs-7). However, the pack lacks explicit documentation of test cards for disputes/chargebacks specifically, or a catalog of scenario-specific test cards, and community evidence notes real-world integration friction (adyen-comm-9) though not specifically about sandbox fidelity. Missing for 10: explicit dispute/chargeback simulation test cards, a documented full test-card matrix, and independent hands-on confirmation that sandbox mirrors production behavior.

   • [claimed-docs] “We recommend testing your full integration, including the client-side and server-side parts, in your test environment before accepting live …”
   • [claimed-docs] “Test the different types of transactions with Adyen that your integration handles using test credentials.”
   • [claimed-docs] “When the payment or modification request is unsuccessful, the `resultCode`, `refusalReason`, and `refusalReasonCode` describe the issue.”
   • [claimed-docs] “In your Customer Area, you can find a log of all webhook events sent to your server.”
   • [community] “Longtime Stripe user did about a year-long stint with Adyen. Stripe beats Adyen on ease of integration, ease of using the back office, charg…”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle documents a full Sandbox environment that mirrors production ("same features and APIs but completely separate datasets, credentials, and dashboards") with sandbox-specific API keys, and the MCP tooling explicitly supports simulating webhook events for testing. However, the evidence pack contains no documentation of specific test card numbers for scenario coverage (e.g., decline codes, dispute/chargeback simulation), and notes that Paddle Retain's dunning/payment-recovery features explicitly cannot be tested in sandbox ('you can't integrate or test with sandbox accounts'), which undercuts full parity for testing the entire payment flow. missing for 10: documented test-card list for various decline/failure scenarios, explicit dispute/chargeback simulation support, and confirmation that all payment-flow features (including Retain) are testable in sandbox.

   • [claimed-docs] “Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments.”
   • [claimed-docs] “"Sandbox is a separate Paddle environment for building and testing your integration without affecting real data or accepting real payments..…”
   • [claimed-docs] “Generate integration code, configure webhooks, and simulate events for testing.”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”

   proof ↗⚑ flag

Ai disputes

1. ai-native userHave an agent handle a dispute end-to-end, retrieving it, assembling evidence, and submitting the response via the API

   weight 3 · round drawn
   Adyen—none0/10

   The evidence pack contains no mention of a Disputes API, chargeback evidence submission, or any agent/MCP workflow for handling disputes end-to-end—only risk/fraud detection (Protect), refusal reasons, and unrelated agentic-commerce checkout flows are documented. The MCP server's tool list (sessions, payment links, modifications, management API) explicitly does not include dispute retrieval or evidence submission.

   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “You can detect, monitor, and mitigate fraud using Protect, Adyen's risk management system.”
   • [claimed-docs] “Each payment goes through a risk evaluation. The risk evaluation reaches a decision that results in an action to block, allow, or review the…”
   • [claimed-docs] “When the payment or modification request is unsuccessful, the `resultCode`, `refusalReason`, and `refusalReasonCode` describe the issue.”

   proof ↗⚑ flag

   Paddle—none0/10

   The evidence pack contains no mention of a Disputes/Chargebacks API endpoint, no ability to retrieve dispute details, assemble evidence, or submit a response programmatically. The only related mentions are generic references to 'chargeback defense' as part of Paddle's merchant-of-record fee and 'investigate failed payments' in the MCP server description, neither of which constitutes dispute-evidence submission via API.

   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”

   proof ↗⚑ flag

Disputes

1. finance-ops userRespond to a chargeback with structured evidence and track the dispute lifecycle to resolution

   weight 3 · round drawn
   Adyen—none0/10

   The evidence pack covers Adyen's risk/fraud tooling (Protect, rule builder, ML fraud detection) but contains no documentation of a chargeback/dispute management API, structured evidence submission, or dispute lifecycle tracking/status workflow. A vague community comment says Stripe beats Adyen on chargebacks 'hands down' but gives no concrete detail to substantiate a documented-capability failure, so it doesn't rise to a disputed claim — it simply reinforces the absence of positive evidence.

   ⚑ flag

   Paddle—none0/10

   Evidence only shows chargebacks appearing as a fee line-item in payout reconciliation reports and a passing mention that Paddle's merchant-of-record fee includes 'chargeback defense,' but there is no documentation of a finance-ops workflow for submitting structured dispute evidence or tracking a chargeback's lifecycle to resolution.

   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”
   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”

   proof ↗⚑ flag

Fraud prevention

1. finance-ops userScreen transactions with machine-learning fraud scores and write custom block, review, and allow rules

   weight 3 · round to Adyen
   Adyen✓fullclaimed8/10

   Adyen's Protect risk management system explicitly provides ML-based fraud scoring plus a rule builder to create, backtest, and label custom rules, with risk evaluations resulting in block/allow/review actions — directly matching the story. Missing for 10: independent/hands-on validation of the rule builder UX and finance-ops-specific workflow evidence beyond first-party docs.

   • [claimed-docs] “You can detect, monitor, and mitigate fraud using Protect, Adyen's risk management system.”
   • [claimed-docs] “Create, backtest, and label custom rules with a rule builder interface | Machine learning powered fraud detection using global transaction d…”
   • [claimed-docs] “Machine learning powered fraud detection using global transaction data | | | Create, backtest, and label custom rules with a rule builder in…”
   • [claimed-docs] “Each payment goes through a risk evaluation. The risk evaluation reaches a decision that results in an action to block, allow, or review the…”
   • [claimed-docs] “Adyen Uplift is the platform's AI optimization layer: "Adyen Uplift helps you strike a balance between conversion, risk and cost. You get re…”

   proof ↗⚑ flag

   Paddle—none0/10

   Evidence only mentions Paddle's merchant-of-record fee covering generic 'fraud protection' and 'chargeback defense' as a bundled service, with no documentation of ML-generated fraud scores or of any UI/API for finance-ops to write custom block/review/allow rules. This axis is fair for a payments platform, but no capability is evidenced.

   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”

   proof ↗⚑ flag

Refunds

1. finance-ops userIssue full or partial refunds from both the dashboard and the API

   weight 3 · round to Adyen
   Adyen~partialclaimed5/10

   Evidence confirms API-level refund capability (Adyen's Management/Checkout 'modifications' API allows refunding a captured payment, and the MCP tool set explicitly includes 'refund a captured payment') and dashboard access to transaction management via the Customer Area, but no citation explicitly confirms partial-refund amount support or a documented dashboard refund button/workflow. missing for 10: explicit documentation of partial refund support (specifying an amount less than the captured total), explicit dashboard refund action screenshot/flow, and any independent/hands-on confirmation that refunds work as described.

   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “When the payment or modification request is unsuccessful, the `resultCode`, `refusalReason`, and `refusalReasonCode` describe the issue.”
   • [claimed-docs] “Manage your account and transactions from our Customer Area, which shows an overview of all your payment channels.”
   • [claimed-docs] “In your Customer Area, you can find a log of all webhook events sent to your server.”

   proof ↗⚑ flag

   Paddle~partialclaimed4/10

   Refunds are mentioned only indirectly, e.g. AI/MCP docs say agents can 'process customer refunds' and MCP can 'analyze revenue, refunds, and transaction patterns,' implying refund capability exists in the underlying API, but no dedicated dashboard or API documentation on issuing full/partial refunds is present in the evidence pack. missing for 10: explicit API endpoint or dashboard walkthrough for issuing refunds, confirmation of partial-refund support, and independent/hands-on evidence of refund flow working.

   • [claimed-docs] “Connect AI agents like Claude and Cursor directly to Paddle to work through complex flows in seconds. Build pricing models, investigate fail…”
   • [claimed-docs] “Build pricing models, investigate failed payments, integrate Paddle features, or process customer refunds — all in natural language.”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “Process subscription upgrades, downgrades, pauses, and cancellatio”

   proof ↗⚑ flag

Multi currency

1. founderCharge customers in their local currencies and settle to my bank in mine with transparent FX handling

   weight 2 · round to Paddle
   Adyen~partialclaimed7/10

   Adyen documents accepting local payment methods/currencies, configuring multiple payout currencies per merchant account, and both local (no SWIFT fee) and cross-border (SWIFT) settlement options, plus flexible payout timing/currency choice — directly supporting charge-local/settle-in-own-currency workflows. However, there's no explicit documentation of FX rate transparency (e.g., disclosed markup, real-time rate reporting) beyond the currency/payout mechanics themselves. Missing for 10: explicit FX rate/markup transparency documentation, independent confirmation of settlement FX handling in practice.

   • [claimed-docs] “Local payouts are faster and avoid SWIFT fees. You need a local bank account in a country that uses that specific currency as the legal tend…”
   • [claimed-docs] “You can configure multiple payout currencies for each merchant account.”
   • [claimed-docs] “Flexible payouts Choose when and in which currency you want to settle.”
   • [claimed-docs] “You can receive payouts in two ways: Local payout ... Cross-border payout: We send funds through SWIFT.”
   • [claimed-docs] “You can configure the payout frequency for each of your merchant accounts.”
   • [claimed-docs] “This means you have the option to change how you get paid for each of your merchant accounts.”

   proof ↗⚑ flag

   Paddle✓fullclaimed8/10

   Paddle explicitly documents built-in multi-currency support for both payments and payouts ('customers purchase in their local currency and letting you get paid in yours'), plus localized pricing pages, 200+ country/currency support, and payout reconciliation reports that break down amounts in both transaction and balance currencies for transparency. As merchant of record, Paddle also handles tax compliance alongside FX, directly matching the story. Missing for 10: no explicit documentation of FX conversion rates/margins used, and no independent/hands-on corroboration of settlement transparency beyond vendor docs.

   • [claimed-docs] “Paddle has built-in support for multi-currency for both payments and payouts, letting customers purchase in their local currency and letting…”
   • [claimed-docs] “you can unlock new revenue by selling in over 200 countries and territories across the world — no additional setup needed.”
   • [claimed-docs] “Build pricing pages that show customers prices localized for their country and currency — including estimated taxes and discounts.”
   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”
   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”
   • [claimed-docs] “Available in over 17 languages, across 200+ markets.”

   proof ↗⚑ flag

Tax

1. finance-ops userCalculate and collect the correct sales tax, VAT, or GST on every transaction automatically

   weight 3 · round to Paddle
   Adyen—none0/10

   Adyen's evidence pack is entirely about payment processing, payouts, platforms, risk management, and agentic-commerce integrations — nothing addresses tax calculation, VAT/GST determination, tax collection, or tax compliance/remittance features. This axis applies since a payments platform could plausibly integrate tax calculation, but no evidence shows Adyen offers this.

   ⚑ flag

   Paddle!disputedcontradicted6/10

   Paddle's merchant-of-record model is core to its tax story: docs explicitly state Paddle 'calculates, collects, and remits taxes for you' across 200+ countries with no extra engineering, and reconciliation reports break out tax rates/tax mode per transaction for compliance proof. However, a hands-on community report describes phenomenon-level VAT invoice errors (wrong post-Brexit VAT number, then wrong currency amounts) contradicting the 'just works' claim, while another user praises the same EU VAT handling as a key differentiator vs Stripe/Chargebee — a genuine mixed/contradictory record rather than mere skepticism. Missing for 10: independent audits or documented tax-accuracy testing across specific jurisdictions, and resolution/rebuttal of the VAT invoice complaint.

   • [claimed-docs] “As a merchant of record, Paddle calculates, collects, and remits taxes for you.”
   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”
   • [claimed-docs] “Build pricing pages that show customers prices localized for their country and currency — including estimated taxes and discounts.”
   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”
   • [community] “I had contacted Paddle for UK VAT invoice for our Setapp for Teams subscription and their incompetence is phenomenon: first they used an EU …”
   • [community] “I use Paddle for taking payments for a paid Chrome extension and they've been great with me. Big draw was 'tax management' handling EU VAT u…”

   proof ↗⚑ flag

1. ai-native userDo everything through the API that I can do in the UI

   weight 2 · round to Paddle
   Adyen~partialprobed5/10

   Adyen exposes extensive Management/Checkout/Platforms APIs covering payments, onboarding, payouts, webhooks, credentials, terminals, and even an MCP server wrapping many of these APIs, suggesting broad API/UI parity for core payment operations. However, some UI-only surfaces (Customer Area account/transaction overview, visual reporting dashboards) are described only as UI features with no explicit statement that every dashboard/report/action is also API-accessible, and no full OpenAPI spec was reachable at standard docs endpoints (404s), only a separate GitHub repo. Missing for 10: explicit parity claim or evidence that all Customer Area/reporting/dashboard actions have API equivalents, and a verified complete OpenAPI coverage confirmation.

   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”
   • [claimed-docs] “Manage your account and transactions from our Customer Area, which shows an overview of all your payment channels.”
   • [claimed-docs] “You can also create multiple API credentials to improve security and control access.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”
   • [claimed-docs] “Adyen API credentials are scoped by construction: each credential carries "Roles: Permissions that define what the credential is allowed to …”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.adyen.com/openapi.json, https://docs.adyen.com/swagger.json, https://docs.adyen.com/api…”
   • [claimed-docs] “The Adyen OpenAPI Specification repository (github.com/Adyen/adyen-openapi) "contains OpenAPI specifications for every public API Adyen offe…”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle's docs show broad API coverage for most UI-surfaced flows — checkout, transactions, subscriptions (pause/upgrade), invoices, webhooks, reports, customer portal — plus MCP/agent-skill tooling that lets agents perform catalog, billing, and integration tasks end-to-end (paddle-docs-3, paddle-docs-15-19, paddle-supp-mcp-server, paddle-supp-agent-skills). However, the pack also documents explicit exceptions, e.g. Paddle Retain's dunning/payment-recovery features 'work with live data... you can't integrate or test with sandbox accounts' (paddle-supp-dunning), showing at least one dashboard-configurable feature not fully API/sandbox-equivalent, and there's no blanket claim or independent audit confirming total UI/API parity. missing for 10: an explicit vendor or third-party statement of full UI-API feature parity, and confirmation that Retain-style dashboard-only configurations have API equivalents.

   • [claimed-docs] “you may also create your own transactions using the API or Paddle dashboard.”
   • [claimed-docs] “Upgrade a subscription Add or remove recurring addons, move from month to annual, or switch plans by”
   • [claimed-docs] “Your sales team can draft and send invoices for subscriptions, collecting payment manually rather than by charging a card on file.”
   • [claimed-docs] “Generate reports via API to feed into accounting and BI systems so monthly reconciliations, checks, and dashboards”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “"Agent skills are a standardized way to give AI agents instructions on how to integrate Paddle... written specifically for agents, rather th…”
   • [claimed-docs] “Failed-payment recovery is built in at two tiers: "If you use Paddle Billing without integrating with Paddle Retain, failed payments for aut…”

   proof ↗⚑ flag

2. ai-native userExport all of my data in open formats and leave

   weight 3 · round drawn
   Adyen—none0/10

   The evidence pack shows reporting dashboards and general documentation transparency (llms.txt, markdown docs, OpenAPI specs) but nothing about a merchant being able to export their own transactional/account data in open formats or close their account and leave without lock-in. Reporting is described as visual dashboards (adyen-docs-10), not bulk open-format export, and no data-portability/account-closure mechanism is documented.

   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”
   • [community] “Longtime Stripe user did about a year-long stint with Adyen. Stripe beats Adyen on ease of integration, ease of using the back office, charg…”

   proof ↗⚑ flag

   Paddle—none0/10

   Paddle documents reports/reconciliation exports and an OpenAPI spec for building integrations, but there is no evidence of a customer-facing 'export all my account/business data in an open format to leave' capability — no GDPR-style data export, no full account data dump, nothing enabling migration away from Paddle with your data intact. Community comments even raise concerns about Paddle's control over customer data rather than showing an exit/export path.

   • [claimed-docs] “Generate reports via API to feed into accounting and BI systems so monthly reconciliations, checks, and dashboards”
   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”
   • [community] “I'm not sure if that pro Paddle list is good for me as a customer. The access to my data at least isn't.”
   • [community] “As a customer, I'd deliberately avoid Paddle to keep my privacy and avoid spam, referencing Paddle's access to customer emails feature.”

   proof ↗⚑ flag

Ai reconciliation

1. ai-native userHave an agent pull payout and balance-transaction data via the API and reconcile each bank deposit to its underlying charges

   weight 3 · round to Paddle
   Adyen—none0/10

   The evidence pack describes payouts, payout frequency, currencies, and a dashboard 'visual representation' of balances/payouts (adyen-docs-9,10,12,30,43), but nowhere documents an API for pulling balance/transaction-level data or a reconciliation workflow that an agent could invoke — the MCP tool list (adyen-supp-mcp-tools) covers checkout, payment links, and management API objects (merchants, terminals, webhooks, credentials) with no balance-transaction or reconciliation tools. Missing for 10: documented Balance/Transaction API endpoints, evidence of agent/MCP access to payout reconciliation data, and any reconciliation workflow example.

   • [claimed-docs] “Adyen supports paying out to cards instantly, and immediately returns the result of the payout attempt.”
   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”
   • [claimed-docs] “You can configure the payout frequency for each of your merchant accounts.”
   • [claimed-docs] “Flexible payouts Choose when and in which currency you want to settle.”
   • [claimed-docs] “You can receive payouts in two ways: Local payout ... Cross-border payout: We send funds through SWIFT.”
   • [claimed-docs] “The Adyen MCP server (github.com/Adyen/adyen-mcp, Alpha) runs locally via `npx -y @adyen/mcp --env=TEST` (or `--env=LIVE --livePrefix=YOUR_P…”

   proof ↗⚑ flag

   Paddle✓fullclaimed7/10

   Paddle publishes dedicated payout reconciliation reports that explicitly reconcile payouts to transactions/adjustments, itemized by gross, tax, fees and currency, generated via API for feeding into accounting/BI systems, and the hosted Paddle MCP server exposes this data to agents (report generation, revenue/refund/transaction analysis) alongside Bearer-token API access with granular permissions. This directly matches the ai-native reconciliation workflow described in the story. missing for 10: independent/hands-on evidence of an agent actually executing an end-to-end payout-to-charge reconciliation via the API/MCP, and explicit documentation of a 'balance transaction' endpoint terminology matching the story's phrasing

   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”
   • [claimed-docs] “Generate reports via API to feed into accounting and BI systems so monthly reconciliations, checks, and dashboards”
   • [claimed-docs] “Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculate…”
   • [claimed-docs] “Generate detailed reports that connect your Paddle payouts to the underlying transaction movements for accurate financial reconciliation.”
   • [claimed-docs] “"The Paddle MCP server is a remote, hosted MCP server" with per-environment endpoints — Sandbox `https://sandbox-mcp.paddle.com/mcp` (API ke…”
   • [claimed-docs] “Bearer-token API keys with documented granular permissions: keys are created in "Paddle > Developer tools > Authentication", scoped per the …”
   • [claimed-docs] “Use Bearer authentication when making requests to the Paddle API. Get a key from Paddle > Developer tools > Authentication.”

   proof ↗⚑ flag

Payouts

1. finance-ops userControl the payout schedule and see exactly which charges, fees, and refunds make up each payout

   weight 2 · round drawn
   Adyen~partialclaimed6/10

   Docs clearly support configuring payout frequency, currency, and timing per merchant account (adyen-docs-12, adyen-docs-30, adyen-docs-50, adyen-docs-22) and note local vs cross-border payout mechanics (adyen-docs-13, adyen-docs-43). Reporting is mentioned as giving a 'visual representation of balances, sales, and payouts' (adyen-docs-10) and the Customer Area gives an overview of payment channels (adyen-docs-16), but there is no explicit evidence of a line-item payout reconciliation report breaking down which specific charges, fees, and refunds compose each payout. Missing for 10: detailed payout reconciliation report/API showing per-payout transaction breakdown, fee itemization, and refund matching.

   • [claimed-docs] “You can configure the payout frequency for each of your merchant accounts.”
   • [claimed-docs] “Flexible payouts Choose when and in which currency you want to settle.”
   • [claimed-docs] “This means you have the option to change how you get paid for each of your merchant accounts.”
   • [claimed-docs] “You can configure multiple payout currencies for each merchant account.”
   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”
   • [claimed-docs] “Manage your account and transactions from our Customer Area, which shows an overview of all your payment channels.”
   • [claimed-docs] “Local payouts are faster and avoid SWIFT fees. You need a local bank account in a country that uses that specific currency as the legal tend…”
   • [claimed-docs] “You can receive payouts in two ways: Local payout ... Cross-border payout: We send funds through SWIFT.”

   proof ↗⚑ flag

   Paddle~partialclaimed6/10

   Paddle provides detailed payout reconciliation reports that itemize gross, taxes, Paddle fees, chargebacks, and retained fees per payout, in both transaction and balance currencies, split by invoice entity — directly matching the 'see exactly which charges/fees/refunds' half of the story. However, no evidence describes finance-ops controlling or configuring the payout schedule itself (frequency, timing, thresholds). Missing for 10: documentation of payout schedule/frequency configuration, independent/hands-on confirmation of reconciliation report accuracy.

   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”
   • [claimed-docs] “Generate reports via API to feed into accounting and BI systems so monthly reconciliations, checks, and dashboards”
   • [claimed-docs] “Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculate…”
   • [claimed-docs] “Generate detailed reports that connect your Paddle payouts to the underlying transaction movements for accurate financial reconciliation.”

   proof ↗⚑ flag

Reconciliation

1. finance-ops userReconcile settlements to bank deposits with itemized reports covering fees, refunds, and chargebacks

   weight 3 · round to Paddle
   Adyen~partialclaimed3/10

   Adyen's reporting docs mention a dashboard 'visual representation of your balances, sales, and payouts' and configurable payout frequency/currency, implying some settlement reporting exists, but nothing in the evidence describes itemized reconciliation reports that break out fees, refunds, and chargebacks against bank deposits. Missing for 10: documented settlement/reconciliation report format, explicit fee/refund/chargeback itemization, and any bank-deposit matching workflow or independent confirmation it works as needed.

   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”
   • [claimed-docs] “You can configure the payout frequency for each of your merchant accounts.”
   • [claimed-docs] “Manage your account and transactions from our Customer Area, which shows an overview of all your payment channels.”
   • [claimed-docs] “You can configure multiple payout currencies for each merchant account.”
   • [claimed-docs] “Flexible payouts Choose when and in which currency you want to settle.”
   • [claimed-docs] “You can receive payouts in two ways: Local payout ... Cross-border payout: We send funds through SWIFT.”

   proof ↗⚑ flag

   Paddle✓fullclaimed8/10

   Paddle publishes dedicated payout reconciliation documentation describing itemized breakdowns of gross, taxes, fees, chargebacks and retained fees split by transaction/balance currency, invoice entity, and even US state/ZIP, explicitly framed as reconciling payouts to remittance advice/bank deposits, plus an API to feed accounting/BI systems for monthly reconciliation. missing for 10: independent/hands-on customer confirmation that the reports actually match bank deposits in practice, and no detail on report formats/exports beyond API generation.

   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”
   • [claimed-docs] “Generate reports via API to feed into accounting and BI systems so monthly reconciliations, checks, and dashboards”
   • [claimed-docs] “Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculate…”
   • [claimed-docs] “Generate detailed reports that connect your Paddle payouts to the underlying transaction movements for accurate financial reconciliation.”
   • [claimed-docs] “Paddle has built-in support for multi-currency for both payments and payouts, letting customers purchase in their local currency and letting…”

   proof ↗⚑ flag

Marketplace

1. developerBuild a marketplace or platform that onboards sellers, splits payments between parties, and pays out to each seller

   weight 2 · round to Adyen
   Adyen✓fullclaimed8/10

   Adyen for Platforms is explicitly documented to onboard sellers/contractors with verification (adyen-docs-11, adyen-docs-31, adyen-docs-36), split payments between parties and hold funds until payout (adyen-docs-21, adyen-docs-37), and configure flexible payout frequency/currency/local vs cross-border payouts (adyen-docs-12, adyen-docs-13, adyen-docs-22, adyen-docs-43, adyen-docs-50). Reporting/dashboards and risk management further round out the marketplace platform story (adyen-docs-10, adyen-docs-8). missing for 10: independent/hands-on developer account of building a full marketplace integration end-to-end, and more granular API-level detail (e.g., specific split-payment API request/response schemas) beyond docs summaries.

   • [claimed-docs] “Onboard sellers, service providers, or contractors on to your platform, and let Adyen verify them before paying out.”
   • [claimed-docs] “Process payments: accept payments on behalf of your users using Adyen's payment processing platform. You can split the payments between one …”
   • [claimed-docs] “Onboard sellers, service providers, or contractors on to your platform, and let Adyen verify them before paying out. Adyen informs you of th…”
   • [claimed-docs] “Onboard and verify your users: onboard sellers, service providers, or contractors on to your platform, and let Adyen verify them before payi…”
   • [claimed-docs] “You can split the payments between one or more users, deduct costs as needed, and hold funds until payout.”
   • [claimed-docs] “You can configure the payout frequency for each of your merchant accounts.”
   • [claimed-docs] “Local payouts are faster and avoid SWIFT fees. You need a local bank account in a country that uses that specific currency as the legal tend…”
   • [claimed-docs] “You can configure multiple payout currencies for each merchant account.”
   • [claimed-docs] “You can receive payouts in two ways: Local payout ... Cross-border payout: We send funds through SWIFT.”
   • [claimed-docs] “Allow your users to easily sign up, sell, and receive payments all within one solution.”
   • [claimed-docs] “This means you have the option to change how you get paid for each of your merchant accounts.”
   • [claimed-docs] “Get a visual representation of your balances, sales, and payouts.”

   proof ↗⚑ flag

   Paddle—none0/10

   The evidence pack describes Paddle purely as a merchant-of-record billing/subscription platform for a single seller (checkout, subscriptions, invoices, payouts to the seller's own account, tax compliance) — there is no mention of onboarding multiple third-party sellers, split payments between parties, or marketplace payout distribution akin to Stripe Connect. Nothing in the docs, MCP tooling, or SDKs addresses multi-party marketplace flows.

   • [claimed-docs] “Paddle has built-in support for multi-currency for both payments and payouts, letting customers purchase in their local currency and letting…”
   • [claimed-docs] “With Paddle, you can unlock new revenue by selling in over 200 countries and territories across the world — no additional setup needed.”
   • [claimed-docs] “As a merchant of record, Paddle calculates, collects, and remits taxes for you.”
   • [claimed-docs] “The merchant-of-record model is the tax story: "Go global and sell in over 200 countries and territories, fully tax compliant, with no extra…”
   • [claimed-docs] “"Payout reconciliation reports help you reconcile payouts to transactions and adjustments to understand how your payout amounts are calculat…”

   proof ↗⚑ flag

1. ai-native userChoose where my data is stored (region/residency)

   weight 2 · round drawn
   Adyen—none0/10

   The evidence pack contains no mention of data residency, regional data storage options, or any control over where Adyen stores payment/customer data. While this is a fair question for a payments platform (compliance/data-residency is a common enterprise requirement), nothing in the docs, community posts, or agentic-commerce material addresses it.

   ⚑ flag

   Paddle—none0/10

   No evidence addresses data residency or region-specific storage choices for Paddle accounts/customers; the pack covers billing, checkout, tax compliance, and AI/MCP tooling but never data residency or regional storage options. Missing for 10: any mention of data center regions, data residency guarantees, or ability to select storage location.

   ⚑ flag

2. ai-native userControl data retention and deletion

   weight 2 · round drawn
   Adyen—none0/10

   Adyen's evidence pack covers payments, payouts, risk management, and agentic commerce integrations, but there is no documentation on data retention policies, data deletion requests, or user/API-level controls to manage how long payment or transaction data is stored or how to purge it. This is a fair axis for a payments platform handling PII/PCI data, but no evidence addresses it.

   ⚑ flag

   Paddle—none0/10

   The evidence pack contains extensive docs on billing, checkout, webhooks, API auth, and AI/MCP integrations, but nothing about data retention policies, customer data deletion/export controls, or GDPR-style right-to-be-forgotten mechanisms. Community comments even raise unresolved privacy concerns about Paddle's access to customer data (paddle-comm-3, paddle-comm-4) without any documented deletion/retention controls to address them.

   • [community] “I'm not sure if that pro Paddle list is good for me as a customer. The access to my data at least isn't.”
   • [community] “As a customer, I'd deliberately avoid Paddle to keep my privacy and avoid spam, referencing Paddle's access to customer emails feature.”

   proof ↗⚑ flag

3. ai-native userOpt out of telemetry and usage tracking

   weight 2 · round drawn
   Adyen—none0/10

   Adyen ships AI-native tooling (an MCP server, docs-for-LLMs surface, agentic commerce APIs) so a privacy-posture question about telemetry/usage-tracking opt-out is a fair axis to ask, but no evidence in the pack describes any telemetry collection by these tools or any opt-out mechanism for it.

   ⚑ flag

   Paddle—none0/10

   The evidence pack contains no mention of telemetry or usage-tracking collection by Paddle's SDKs, MCP servers, or dashboard, nor any opt-out mechanism. Community items about customer data/spam concerns (paddle-comm-3, paddle-comm-4) relate to end-customer marketing data, not AI-agent telemetry, so they don't establish or contradict this capability.

   ⚑ flag