generated: '2026-09-19' method: probed source: https://api.berrergate.com/.well-known/agent-card.json card: file: a2a/berrergate-com-agent-card.json name: BerrerGate Tool & Provider Intelligence url: https://api.berrergate.com/a2a/jsonrpc version: 2.9.0 skills: 12 discovery: path: /.well-known/agent-card.json canonical: true host: api.berrergate.com note: >- Served from the API host, which is also the OpenAPI servers[] host, the MCP host and both A2A endpoint hosts. The apex berrergate.com 302s the canonical path to this URL, and the homepage's primary call to action ("Agent Card") links straight to it. The legacy /.well-known/agent.json on the same host serves the identical 8,039-byte body. Unknown /.well-known/* paths return the service's real JSON 404 and a negative-control path also 404s, so the 200 is a served document. Ownership is not in question: the card's provider.organization is "Berrer" with provider.url https://berrergate.com, the OpenAPI on the same host titles itself "BerrerGate Tool & Provider Intelligence API", the homepage JSON-LD names the provider Organization "Berrer", and the Global A2A Registry lists the agent under Berrer as Verified. conformance: spec: A2A 1.0.0 grade: flavored grade_note: >- Graded against the catalog's three hard checks (lib_agent_card.grade_agent_card): capabilities IS an object and skills IS an array, but there is NO top-level protocolVersion — the card carries protocolVersion "1.0" inside each supportedInterfaces[] entry (the proto-derived 1.0 shape: AgentInterface{url, protocolBinding, protocolVersion}) and no top-level url, preferredTransport or additionalInterfaces. By the rubric that is a hard-check failure (no-protocolVersion) plus the supportedInterfaces deviation, so the grade is flavored; a reader should know the card is nonetheless internally consistent with the 1.0 interface model and was accepted as Verified by the Global A2A Registry's own validator. It also omits the optional preferredTransport, and adds non-schema fields (category, target_audience, extensions.berrergate). protocol_version: '1.0' protocol_version_location: supportedInterfaces[].protocolVersion (not top-level) preferred_transport: null interfaces: - {url: 'https://api.berrergate.com/a2a/v1', protocolBinding: HTTP+JSON, protocolVersion: '1.0'} - {url: 'https://api.berrergate.com/a2a/jsonrpc', protocolBinding: JSONRPC, protocolVersion: '1.0'} deviations: - no-protocolVersion - supportedInterfaces-instead-of-additionalInterfaces - no-top-level-url - no-preferredTransport - 'non-schema-fields (category, target_audience, extensions.berrergate)' - 'securitySchemes-empty-object-and-securityRequirements-empty (anonymous by design; paid skills are x402-gated over HTTP)' passes: - 'capabilities is an object ({streaming: false, pushNotifications: false, extendedAgentCard: false})' - skills is an array of 12 objects each with id, name, description, tags, examples, inputModes, outputModes - 'defaultInputModes [application/json, text/plain] and defaultOutputModes [application/json] present' - 'provider {organization, url} and documentationUrl present' - served with content-type application/a2a+json x-evidence: fetched: '2026-09-19' url: https://api.berrergate.com/.well-known/agent-card.json http_status: 200 content_type: application/a2a+json; charset=utf-8 body_bytes: 8039 body_parses_as: JSON object with AgentCard shape (name, description, supportedInterfaces, provider, version, documentationUrl, capabilities, securitySchemes, securityRequirements, defaultInputModes, defaultOutputModes, skills, extensions) corroborating_probes: - url: https://api.berrergate.com/.well-known/agent.json http_status: 200 note: Legacy path; identical body. - url: https://berrergate.com/.well-known/agent-card.json http_status: 302 note: Redirects to the API-host card. - url: https://www.berrergate.com/.well-known/agent-card.json http_status: 522 note: Cloudflare origin unreachable on the www alias for every path. - url: https://api.berrergate.com/a2a/jsonrpc method: POST body: '{"jsonrpc":"2.0","id":1,"method":"tasks/get","params":{"id":"apievangelist-nonexistent-probe"}}' http_status: 200 response: '{"jsonrpc":"2.0","id":1,"error":{"code":-32601,"message":"Method not found"}}' note: >- A JSON-RPC 2.0 responder that does NOT implement tasks/get — consistent with the OpenAPI summary "A2A JSON-RPC v1.0 SendMessage compatibility interface" (message/send only). No message was sent and nothing was purchased. - url: https://api.berrergate.com/a2a/jsonrpc method: POST body: '{"jsonrpc":"2.0","id":2,"method":"agent/getAuthenticatedExtendedCard","params":{}}' http_status: 200 response: '{"jsonrpc":"2.0","id":2,"error":{"code":-32601,"message":"Method not found"}}' - url: https://api.berrergate.com/a2a/v1 http_status: 404 note: GET on the HTTP+JSON interface root returns the JSON 404; the declared route is POST /a2a/v1/message:send. - url: https://api.berrergate.com/.well-known/apievangelist-negative-control-7f3a9c.json http_status: 404 - url: https://www.a2a-registry.org/agent/com.berrergate.berrergate_capability_wisdom_agent http_status: 200 note: >- The Global A2A Registry listing that put this provider in the harvest backlog: id com.berrergate.berrergate_capability_wisdom_agent, owner Berrer, "Registered: Aug 27, 2026", badge "Verified Agent", and — at fetch time — "Unreachable: This agent has failed 14 consecutive health checks", which contradicts our own same-day 200s on the card, the OpenAPI and the JSON-RPC endpoint. The registry was the lead; the card above was fetched directly from the provider's host. agent_card: name: BerrerGate Tool & Provider Intelligence description: >- Agent-first capability intelligence with free discovery, a wisdom-first paid Evidence Brief, low-cost no-web inference, and commercially cleared stack decisions. Payments use Base USDC via x402. version: 2.9.0 provider: {organization: Berrer, url: https://berrergate.com} documentation_url: https://api.berrergate.com/openapi.json capabilities: {streaming: false, pushNotifications: false, extendedAgentCard: false} security_schemes: {} security_requirements: [] default_input_modes: [application/json, text/plain] default_output_modes: [application/json] skill_count: 12 skills: - {id: berrergate-find-agent-tool-api-provider, name: 'Find an Agent Tool, API or Provider', cost: free} - {id: berrergate-procurement-preview, name: Get a Free Procurement Preview, cost: free} - {id: berrergate-spend-provider-preview, name: 'Scout Gift Card, Top-Up & eSIM Providers', cost: 'free (read-only preview)'} - {id: berrergate-zero-signup-commerce-discovery, name: 'Discover Digital Goods, x402 Services, MCP Servers & Public APIs', cost: 'free (cache-only)'} - {id: berrergate-evidence-brief, name: Berrer Evidence Brief, cost: '$0.020 USDC via x402 on Base'} - {id: berrergate-micro-inference, name: Low-Cost Pay-Per-Call AI Answer, cost: 'about $0.00125-$0.00225 USDC, dynamic'} - {id: berrergate-purchase, name: Buy a Commercially Cleared Open-Source Stack Decision, cost: 'purchase instructions for $0.01 packets'} - {id: berrergate-paid-1-agent-agent-observability, name: Choose an Open-Source Agent Observability Stack, cost: '$0.01 USDC'} - {id: berrergate-paid-2-agent-automation-orchestration, name: Choose an Open-Source Automation Orchestration Stack, cost: '$0.01 USDC'} - {id: berrergate-paid-3-agent-browser-scraping-automation, name: Choose an Open-Source Browser & Scraping Stack, cost: '$0.01 USDC'} - {id: berrergate-paid-4-agent-product-analytics, name: Choose an Open-Source Product Analytics Stack, cost: '$0.01 USDC'} - {id: berrergate-paid-5-agent-browser-automation, name: Choose an Open-Source Browser Automation Stack, cost: '$0.01 USDC'} extensions_note: >- extensions.berrergate is a provider-specific object of release flags (release_contract, evidence-brief price 20000 atomic, purchase_authority false, active_utility_keys [inference]); it is not an A2A AgentExtension declaration and no capabilities.extensions[] (e.g. a2a-x402) is declared even though the paid skills settle over x402.