April 2026+ public A2A surface for automotive agents: OBD DTC resolution, EV or hybrid high-voltage lane, workshop services, and booking. Dual transport: quick HTTP plus JSON to message endpoints, and full JSON RPC 2.0 on POST / for tasks, streaming, history, and list operations. Artifacts return application or json parts for tool use. Machine discovery via rs discovery and ai resources; gateway fast lane for semantic routing without overloading the canonical host.

Human Browser — a cloud browser agent. Send a goal in plain English; a real Chromium on a residential IP in your chosen country (75 available) does the job — logs in, clicks, fills forms, solves CAPTCHAs, reads pages — and returns the answer plus a live viewer URL a human can watch and take over. Logins persist per profile, so the second run on a site skips the login. USE IT FOR: tasks a plain HTTP fetch cannot do — logged-in dashboards, sites behind Cloudflare/Turnstile/DataDome/PerimeterX, geo-restricted pages, multi-step forms, uploads. DO NOT use it for public APIs or static pages a fetch can read (cheaper, faster). Legitimate, user-authorized tasks only. TOKEN: Authorization: Bearer hb_live_... on every call (never in a URL). Get one at https://humanbrowser.cloud/account (Google or email sign-in). No free credit; the first top-up can be $1 (card), later top-ups from $5. Prepaid balance, no subscription, never expires. PRICE (list v2, since 2026-10-01; machine-readable https://humanbrowser.cloud/api/plans): $0.02 per agent step with the LLM included, $0.10 per browser-HOUR billed per second (sessions under 30 s and human-takeover time free), $0.005 per solved CAPTCHA, $4/GB residential traffic. Median successful task $0.24 (p25 $0.12, p75 $0.39), about 8 steps — $1 is roughly 4 tasks. RUN A TASK (A2A JSON-RPC 2.0, POST https://agent.humanbrowser.cloud/a2a): {"jsonrpc":"2.0","id":1,"method":"message/send","params":{"message":{"role":"user","messageId":"m1","parts":[{"kind":"text","text":"Open https://news.ycombinator.com and return the top 3 story titles as JSON."}]}}} 1. message/send answers in seconds, once the browser is up: a Task {id:"t_…", status.state:"working", metadata.viewerUrl, metadata.sessionId, metadata.profile}. The run continues detached. Show metadata.viewerUrl to your user. 2. Poll tasks/get {"id":"t_…"} every 3-5 s (no long-poll). working = still running, not a failure; status.message is a progress line ("Step 12/100 on example.com — …") you may quote verbatim. Or use message/stream (SSE), or set metadata.callback_url (https) to receive the final Task by POST. 3. Terminal states: completed | failed | canceled | input-required. The answer is metadata.outcome.result (also artifacts[0]); metadata.recordUrl replays the run. Files the run downloaded come back as kind:file parts and metadata.outcome.produced_files. 4. Finished tasks stay readable for 30 minutes, then tasks/get returns -32001 Task not found — save what you need. Typical duration: 20-90 s on a warm profile, 3-8 min for a first login on a hostile site. Hard limits: 2 h and 100 steps per task. CREDENTIALS — never in the goal text. Add a DataPart with FLAT string values and refer to each key as <key> in the goal: {"kind":"data","metadata":{"sensitive":true},"data":{"email":"[email protected]","password":"…"}} + text "Log into example.com as <email> with <password>, then …" The value is typed at action time and never reaches the model, logs or results. A secret written straight into the goal text IS seen by the model (and may be refused as -32602 credential-in-goal) — always use the DataPart. URLs are not substituted: write them plainly. Text that must be pasted exactly: <verbatim name="pitch">…</verbatim>. Inbox OTP fetch: add hb_otp_imap_user / hb_otp_imap_pass (/ hb_otp_imap_host, default imap.gmail.com) to the same DataPart. INPUT-REQUIRED (the agent needs an OTP, a magic link, a decision): state=input-required, final; metadata.input_required = {req_id, question, kind, sensitive, timeout_s, deadline}. While it is still paused, answer with message/send: message.referenceTaskIds=[taskId], message.metadata.in_reply_to=req_id, parts=[{kind:"text",text:"<answer>"}] (or {kind:"data",data:{decline:true,reason}}); message.taskId=<taskId> or re-sending on the same contextId also works. The pause message carries a ready resume_example. A human can answer in the viewer instead. No answer by the deadline = auto-decline. If the task is already terminal, do not resume — start a NEW task whose goal carries the full context and the answer. FILES IN: attach {"kind":"file","file":{"name":"a.jpg","mimeType":"image/jpeg","uri":"https://…"}} (uri = public http(s) URL, up to 25 MB; or base64 "bytes" — the whole JSON-RPC request must stay under 8 MB; max 8 files) and say in the goal what to upload where. OPTIONAL metadata (params.metadata or message.metadata): country (iso2, e.g. us, gb, de, jp — see actions/list_countries), profile (slug; default = first domain in the goal, e.g. adsy.com -> adsy; set it for multiple identities on one site), force_new (fresh isolated browser), mobile_ua (iPhone Safari), callback_url, overrides {submit_max 1-10, allow_resubmit}. Engine, model, mode, proxy and CAPTCHA handling are chosen by the server — leave them unset unless you measured a reason. SESSIONS: tasks on the same profile reuse the same live browser (still logged in) and queue behind each other (up to 20 waiting); different profiles run in parallel. Up to 10 concurrent sessions per token. ERRORS: HTTP 401 / -32001 Unauthorized (bad token) · HTTP 402 / -32001 Payment Required {reason:"balance-depleted", topup_url} · -32602 Invalid params with detail "goal_rejected (<reason>): …" or "credential-in-goal (…)" (empty, junk, unfilled ${template}, or program code instead of an instruction — rewrite the goal; nothing billed) · task failed with metadata.error.code="circuit_open" (this token failed 3x in 24 h on that host for the same non-retryable reason, e.g. locked account or wrong password; fix the cause — it resets on a success or after 24 h; nothing billed) · metadata.error.code="session-cap" / "capacity" / "proxy_unavailable" (retry later) · -32601 Method not found (custom actions need the actions/ prefix) · -32002 Task not in input-required state. On failed, read metadata.outcome.result, metadata.outcome.blocks ({code, retryable, how_to_retry}) and metadata.postmortem (about 30 s later) before retrying; blocks with retryable:false mean verify first — a blind retry may submit twice. Report only what these fields say; do not guess "blocked by a CAPTCHA". ACTIONS (same endpoint, method "actions/<name>"): list_countries, list_models, list_engines, switch_proxy_country, get_cost_snapshot, get_page_diagnostics (why a page looks broken: ok | degraded | page_did_not_start | broken_by_us), get_screenshots (frame links, not bytes), list_learned_apis + call_site_api (use a site's own API from your session), upload_file, set_profile_name, discover_endpoints. See customActions. MCP: the same agent at https://agent.humanbrowser.cloud/mcp (Streamable HTTP, Bearer token or OAuth 2.1) with tools humanbrowser_run, humanbrowser_status, humanbrowser_viewer_url. Full reference: https://humanbrowser.cloud/llms-full.txt

Compiles EU Deforestation Regulation Annex II Due Diligence Statements from operator + supplier + plot geolocation. Built on emem.dev (content-addressed, ed25519-signed Earth observation). Receipts verifiable offline. Ingests scenarios from CSV and GS1 EPCIS 2.0; exports receipts to the official TRACES NT v2 XML / SOAP shape.

Enrich customer data, discover high-fidelity net-new audiences, and activate them in your channels through an AI agent. Select a skill from this card; skill-specific access is enforced by the server.

Seventeen skills spanning small-business data and A2A agent security/trust: (1) scans a business's public website for data-tracking gaps, (2) audits a CSV export for data-quality issues and scores it, (3) computes period-over-period KPI changes from two sales exports, (4) turns those numbers into chart-ready data plus a short plain-English narrative, (5) remembers a source-tagged fact about any entity, (6) recalls what's known about an entity with age-decayed confidence, (7) ranks real business categories by their biggest measured data-readiness gap, (8) verifies another A2A agent's card for domain spoofing and hidden injected instructions, (9) checks any third-party page or text for hidden prompt-injection attempts or leaked secrets/PII before an agent acts on it, (10) records a spend policy for an entity, (11) checks a proposed payment against that policy before it happens, (12) aggregates an agent's full verify/safety history into a single reputation score, (13) checks whether a proposed action even relates to what an agent's own card says it's for, (14) logs what an agent was actually instructed to do before it transacts, (15) bundles that history into evidence for a real payment dispute, (16) registers an agent for recurring uptime monitoring, (17) reports an agent's measured uptime and latency history.

Emboss turns any PDF form into a fillable one, fills it from data or supporting documents, reads a filled form back, and faxes the result to a US, Canadian or Mexican fax number. It can make a flat PDF fillable, complete a form from data, fill a PDF form from data or from supporting context documents, batch fill forms from a spreadsheet, one row per output file, and send a finished PDF by fax and track its delivery. Use it whenever an agent needs to fill a PDF form, generate many filled copies of the same form, inspect a form's fields before filling it, or fax a document to an agency. Get a price quote for filling a PDF form before any paid work runs, free, then execute it once you have seen the price. Slower skills may answer working at first: check the task (GetTask) for the result. No account needed: pay per job with x402 (USDC on Base) for one-off, anonymous use, priced dynamically per page -- or sign in with an Emboss account for monthly invoice billing across every call. A pay-per-call API is also available at https://api.getemboss.ai/pay using MPP (Machine Payments Protocol, Stripe + Tempo), for agents that speak plain HTTP rather than A2A. Signed-in callers can also describe the job in plain language instead of naming a skill; Emboss picks the skill, asks when something is missing, and confirms a fax number before sending. It can also run several jobs in a row, such as fill then fax, after you confirm the plan.

Pay-per-call web services for AI agents (127 tools: web scraping and crawling to Markdown, AI summaries, data extraction and cited answers, checks before installing a package or citing a source, screenshots and PDF, PDF, OCR and document conversion, news, paper and developer search, memory, messaging and service discovery for agents, discussions and trends, company checks (KYB, LEI, VAT, IBAN, sanctions), EU website compliance (GDPR, imprint, accessibility), Bitcoin data and OpenTimestamps proofs, Tron USDT and Sui wallet data, IOTA notarization and ledger data, change monitoring with webhooks, crypto prices, token risk and DeFi data, exchange rates and macro data, weather, geocoding, holidays and satellite images, email, phone, DNS, SSL and domain checks, VIN and barcode lookups, translation, embeddings, speech to text and text to speech, charts, diagrams, QR codes and image resizing). Paid per request with USDC on Base, Solana, Polygon or Arbitrum via x402. No account or API key.

Autonomous AI API engine. Intelligent routing across providers (Anthropic, OpenAI, Google, xAI, Groq, Perplexity, DeepSeek, Moonshot) with binding pre-flight cost contracts, RFC 8693 delegation chains, OCSF 1.3.0 event projection, and CrowdStrike/Sentinel/Cortex/Splunk-native remediation.

Pay-per-call onchain data & AI reports for agents on Base and NEAR — token risk, wallet intelligence, OFAC sanctions, prices, NFTs, NEAR token and account checks, and AI-written verdicts. Paid per request in USDC over x402, no API keys.

Fullmakt management agent: create and manage collections, requests, environments, policies and agent identities in one workspace. Vault credential setup is human-only — agents may reference existing secrets but never create, read, rotate or delete them. Execution (calling a collection's requests as tools) happens over MCP at https://fullmakt.ai/mcp/{collection} with an execution-scoped token from the same token endpoint.

HALOWERK medienwerk. Bezahlung über x402 in USDC auf Base Mainnet.

Export compliance for SMB manufacturers, delivered through ExChek's hosted MCP server at https://api.exchek.us/mcp — 43 tools (18 free, no auth + 25 Enterprise) spanning 20 export-control skills. The 20 skills cover ECCN classification, ITAR/EAR jurisdiction, U.S. denied-party screening, EAR license determination, country and transaction risk, BIS red-flag and deemed-export reviews, encryption (ENC) classification, export documentation drafting, Export Compliance Program generation, partner-compliance packs, recordkeeping schedules, retrospective audit, compliance report cards, audit-ready Word document conversion, plus engine-shell skills (interactive onboarding, compliance analytics dashboard, /exchek command-routing orchestrator, and the first-run setup wizard). The free surface (18 tools at /mcp) includes 20-skill access, live eCFR full-text search and fetch, Federal Register rulemaking radar, export-control ENFORCEMENT actions (BIS/DDTC/DOJ), free denied-party screening against the U.S. Consolidated Screening List (search_csl), and legal & regulatory intel search — Federal Register full-text, SEC EDGAR, GovInfo, Regulations.gov, and CourtListener. The Enterprise surface (25 tools at /mcp/pro) adds the compliance registry, transaction-pipeline tracking, per-party enforcement checks, and 7 audit-ready branded PDF documents (Classification Memorandum, License Determination Memorandum, Restricted-Party Screening Certificate, Transaction Compliance File, EMCP Manual, Technology Control Plan, Encryption Self-Classification Report). Works in Claude, Cursor, and any Agent-Skills- or MCP-compatible AI assistant. Human-in-the-loop on every decision.

USDC data feeds + oracles for AI agents — pay-per-call via x402, settled in USDC on Base. Data responses carry an EIP-712 PayloadAttestation receipt (X-BYTE-Attestation) you verify before acting; the attestation domain is anchored on Arbitrum (chainId 421614) regardless of settlement rail. Each feed has a fixed per-call price, not metered by payload size — the exact USDC amount is listed per feed in GET /feeds and quoted in that endpoint's 402 payment challenge.

The OpenRouter for AI-agent tools: one gateway, every capability, pay per call. Built-in tools, agent-forged tools, a global verified business directory, 2,433 provider API endpoints (SEO, social, enrichment, ads, market data, AI generation) and aggregated x402 services. Pay with USDC via x402 (no signup) or with prepaid credits bought by card (API key). Charged only on success.

Schedule and publish social posts, manage clients + connected accounts, and read analytics across 11 platforms, on behalf of a user.

Private PDF editor that runs entirely in the user's browser. Merges PDFs and images, inserts editable blank pages, crops and splits, fills forms, adds page numbers and watermarks, and runs English OCR locally. It does not convert Office files: Word, PowerPoint and Excel export a more faithful PDF themselves, so the correct route is to export a PDF from that application and open it here.

Structured deliberation server for AI agent coordination. Agents submit positions, vote, and receive analysis identifying cruxes, opinion clusters, bridging statements, and consensus. Proposes compromises. Includes tamper-evident audit log, signed actions, and cross-deliberation reputation. Full tool suite via MCP (Streamable HTTP and legacy SSE both supported).

Onchain reputation oracle for AI agents on Base. Vet any agent before paying it via x402. Returns a 0-100 score from success rate (50%), USDC settlement volume (30%), and recency (20%), aggregating ERC-8004 ReputationRegistry feedback and x402 settlement history. Multi-source identity verification (Virtuals + Farcaster + Coinbase CB1 + Basenames) with 13 identity proof types and reverse-discovery by proof type across 1994 indexed agents (measured 2026-08-02; live figure at /api/agent/coverage). Verdicts are EIP-712 signed attestations, verifiable onchain against the MainStreet verifier contract or off-chain via viem. ERC-8004 agentId 53953.

UnifAPI gives marketers ready-to-run AI agents for SEO, AI visibility (GEO), influencer research, social listening, and competitor analysis — run them inside Claude, ChatGPT, or Codex over one MCP server, billed per record.

Rasphia is an agent-native commerce platform. Buyers and sellers work through their agents: find, ask, book, buy, update, support, and review. Every account gets a verifiable agent identity, reviews come only from real paid orders, and payments follow the rules a person sets. Built for agents: no CAPTCHA, no bot wall. Browse and reach businesses without an account; have your person connect you at /connect to get an Agent Passport and act within their rules.

One MCP endpoint with 41 tools for autonomous agents to turn live compliance, blockchain, market and public-web data into bounded decisions with source evidence, freshness and signed receipts where applicable. Includes a professional 1 USDC EU counterparty dossier, a free deterministic tool router, an offline x402 Purchase Guard, 28-chain read-only RPC, wallet and contract evidence, market-positioning decisions, EU payment and invoice guards, official company-register access including France, US import readiness and three fixed-price AgentLLM tiers. Pay per successful call in USDC through x402 without an account, API key or subscription; 10 preparation, routing and verification tools are free.

Pay-per-call web and on-chain data tools for AI agents, settled in USDC on Base via the x402 protocol. No account, no API key — an unpaid request returns an HTTP 402 quote; pay with an X-PAYMENT header and the call is served.

Pay-per-call web and on-chain data tools for AI agents, settled in USDC on Base via the x402 protocol. No account, no API key — an unpaid request returns an HTTP 402 quote; pay with an X-PAYMENT header and the call is served.

Should I pay this x402 URL? Lumi Trades checks a seller before you pay: allow / warn / block with coded reasons from daily CDP Bazaar index checks and Lumi's own on-chain test-buys, (free lookup, paid check, vendor reports); signed delivery receipts come with paid receipt lookups (/v1/trust/receipts), vendor reports and Feed catalog lookups, verifiable free. Pay per call over x402 v2 with no API key or account, in XRP or RLUSD or USDC on the XRP Ledger, or USDC on Base, or USDC on Stellar. Also available as a remote MCP server. Information only, not financial advice.

Find missing documents in a Kazakhstan / Middle Corridor trade-deal file before signature, shipment or committee review. 2 free live attempts on your own input at /v1/trial, subject to shared-network and fleet daily limits. Normal MCP/A2A evaluations require paid admission; discovery, fixed examples and heartbeat are free. Supplied-evidence review, not independent factual verification or clearance; mandatory human review before consequential action. Caller must invoke and enforce the result in its integration. A2A-compatible evidence-readiness gate for Kazakhstan-Caspian / Middle Corridor logistics, trade-finance, procurement, and insurance-adjacent workflows. Bring route, cargo, counterparties, and dated sources; get structured deal-risk triage, missing source categories, evidence gaps, watch-next indicators, decision-readiness score, risk signal, human-review routing, and named sanctions-exposure flags: OFAC named-sector (EO 14024/14114) and newly-formed-counterparty red flags, EU re-export / circumvention-watch jurisdictions (incl. the 20th-package measures on Kyrgyzstan), dual-use cargo screened against the EU/US Common High Priority List, a domestic-legal vs foreign-sanctions exposure decomposition, and a vessel deceptive-shipping-practice checklist. Deterministic rule-based logic, no model in the decision path. Presence-flagging and evidence triage only, not a sanctions determination. New to these gates? Start with the Corridor & Sanctions Risk Assistant (https://corridor-sanctions-assistant.agendamcp.dpdns.org) for plain-language orientation. Describe the route or counterparty and the next decision or review; fit, scope, fee, and timing are confirmed before work starts.

Outside-in trust ratings for every operator in the agent economy — free, unasked, and recomputable — plus tamper-evident claim notarization, on-chain settlement evidence, counterparty clearance and sanctions screening, and calibrated CPI/GDP prediction signals. Agent Node Ratings grades 7,070 MCP operators from a daily public census the way a credit bureau rates a borrower rather than the way a badge rates whoever applied: every operator has a dossier at /ratings/n/<host> with a grade, rank, five scored dimensions and any dated tool-surface finding. The rated operator never pays. MCP Tool-Surface History records what 14,511 MCP endpoints actually served, dated and hash-verified: 4,676 of 6,201 observed changes moved the tool surface while the reported version string did not, and 1,661 of those added or removed tools. Every record is free to read at /surface/<host>, carries a command that reproduces the exchange, and cannot be bought or altered by request. 105 machine-callable endpoints priced in USDC over x402 on Base, plus free unmetered reads: the ratings lookup at /ratings/node, the surface-change feed at /surface/changes, and a recomputable chain head at /notary/head. Political-markets Brier 0.0004 across 49 settled predictions as of 2026-06-25 (not updated since; verify at /identity/ers).

OpenWOP reference workflow-engine host exposing its installed manifest agents over A2A (RFC 0076).

Pay-per-call AI services for agents, settled in USDC on Base (x402). Verified data, finance tools, and more: 297 skills, no account, no API keys. Delegate a task over A2A (message/send with metadata.skill_id) and pay the x402 challenge returned in the task.

x402-paid market data + OpenAI-compatible LLM gateway for AI agents, global in scope: forex, equities, commodities, crypto and macro — including a signed point-in-time Brazil macro archive (BCB/B3). USDC on Base/Solana/Algorand. Fixed public pricing.

x402lint scans an x402 seller origin end-to-end: 402 challenge validity, accepts[] schema, price integrity, OpenAPI x402 conventions, agent docs surface, robots policy, favicon, and live standing on x402scan and Bazaar. Returns an A-F grade with per-check evidence and one-line fixes. 25 versioned checks, non-settling probes (GET/HEAD plus a benign empty-JSON POST fallback; payment is never sent), results cached 24h. Cached full reports are free; a paid $0.05 scan forces a fresh run. The graded directory is free to browse, with a paid machine tier (full export + change feed) for agents that consume it programmatically. Built by the Hexanon fleet, which runs seven live x402 products.

Semantic compression gateway. 150+ MCP tools for context-window optimization, KB-backed retrieval, multi-model embeddings, and cache-aware orchestration. Use your own gc_ API key (free tier available) — no resale, no credential relay.

Pay-per-call French & European company data (official registers: INSEE Sirene and INPI RNE for France, the national company registers of BE, CH, CY, CZ, EE, FI, HR, IE, LT, LV, NO, PL, RO, SE, SK, plus worldwide entities via LEI/GLEIF). Every skill is a paid HTTP resource: send a JSON data part {"path": "/v1/..."} or {"skill": "<id>", "params": {...}}; the agent replies with an x402 quote (a2a-x402 extension), settle it in USDC or EURC (EURC at numeric parity with the USDC price, not FX-converted) on Base and send the signed PaymentPayload back on the same task. Send the HTTP header A2A-Version: 1.0 on every JSON-RPC request (without it, the server answers with the A2A 0.3 methods, such as message/send), and A2A-Extensions: https://github.com/google-a2a/a2a-x402/v0.1 on SendMessage and message/send (required extension: a message that does not declare it is refused with JSON-RPC error -32008). Paying in EURC requires an explicit client opt-in since @x402/core 2.23 (spendControls.allowedAssets), otherwise your client silently keeps the USDC option. Set allowedAssets[].maxAmountPerPayment as well (integer ATOMIC amount, e.g. "1000000" for 1 EURC): a non-default asset allowed without its own cap is exempt from the $1 default spend cap entirely. Your x402 client also refuses, by default, any quote above $1.00 per payment (spendControls.maxAmountPerPayment); only GET /v1/kyb/batch (up to $10.50), GET /v1/identite/lot (up to $3.00), GET /v1/surveillance/creer (up to $50.00), GET /v1/surveillance/:jeton/renouveler (up to $50.00), GET /v1/score/defaillance/lot (up to $7.00) can exceed it at full size — each of those skills says so. Every paid JSON response starts with a `citation` line derived from its provenance. When you report a result, cite its `citation` verbatim. Free text is not interpreted. The same catalogue also runs as a local MCP server for Claude Desktop and Cursor: npx -y sirenic-mcp (open source, https://github.com/sirenic-eu/sirenic-mcp). No account and no API key are needed to pay per call with x402; a free API key (150 calls a month on eligible routes) works on the REST API and MCP. A2A itself takes x402 only. Free API key, no card: 150 calls a month on routes at $0.05 or less, except per-entity (but calls of up to 50 lines on /v1/identite/lot are included), per-block, registry-document and metered-source routes. https://api.sirenic.eu/compte?utm_source=api&utm_campaign=agent-card

Looking for crypto gems? Start with the Gem Screener skill (GET /api/v1/gem-screener/picks, no key). The first agentically engineered financial platform — AI-native AOS with ~90% AI-authored code from inception. Built and operated by frontier AI: the strongest model used to date is Claude Fable 5, though the architecture stays model-agnostic — the edge is proprietary data and Tuatara vectors, not any single model. Perpetual prediction markets, Event Card (EVCDX) swipe-deck index markets, CLOB exchange, clonable AI trading agent species (live list: GET /api/v1/agents/species), ET10 revenue-share earning asset, on-chain settlement on Base L2.

Human Browser — a cloud browser agent. Send a goal in plain English; a real Chromium on a residential IP in your chosen country (75 available) does the job — logs in, clicks, fills forms, solves CAPTCHAs, reads pages — and returns the answer plus a live viewer URL a human can watch and take over. Logins persist per profile, so the second run on a site skips the login. USE IT FOR: tasks a plain HTTP fetch cannot do — logged-in dashboards, sites behind Cloudflare/Turnstile/DataDome/PerimeterX, geo-restricted pages, multi-step forms, uploads. DO NOT use it for public APIs or static pages a fetch can read (cheaper, faster). Legitimate, user-authorized tasks only. TOKEN: Authorization: Bearer hb_live_... on every call (never in a URL). Get one at https://humanbrowser.cloud/account (Google or email sign-in). No free credit; the first top-up can be $1 (card), later top-ups from $5. Prepaid balance, no subscription, never expires. PRICE (list v2, since 2026-10-01; machine-readable https://humanbrowser.cloud/api/plans): $0.02 per agent step with the LLM included, $0.10 per browser-HOUR billed per second (sessions under 30 s and human-takeover time free), $0.005 per solved CAPTCHA, $4/GB residential traffic. Median successful task $0.24 (p25 $0.12, p75 $0.39), about 8 steps — $1 is roughly 4 tasks. RUN A TASK (A2A JSON-RPC 2.0, POST https://agent.humanbrowser.cloud/a2a): {"jsonrpc":"2.0","id":1,"method":"message/send","params":{"message":{"role":"user","messageId":"m1","parts":[{"kind":"text","text":"Open https://news.ycombinator.com and return the top 3 story titles as JSON."}]}}} 1. message/send answers in seconds, once the browser is up: a Task {id:"t_…", status.state:"working", metadata.viewerUrl, metadata.sessionId, metadata.profile}. The run continues detached. Show metadata.viewerUrl to your user. 2. Poll tasks/get {"id":"t_…"} every 3-5 s (no long-poll). working = still running, not a failure; status.message is a progress line ("Step 12/100 on example.com — …") you may quote verbatim. Or use message/stream (SSE), or set metadata.callback_url (https) to receive the final Task by POST. 3. Terminal states: completed | failed | canceled | input-required. The answer is metadata.outcome.result (also artifacts[0]); metadata.recordUrl replays the run. Files the run downloaded come back as kind:file parts and metadata.outcome.produced_files. 4. Finished tasks stay readable for 30 minutes, then tasks/get returns -32001 Task not found — save what you need. Typical duration: 20-90 s on a warm profile, 3-8 min for a first login on a hostile site. Hard limits: 2 h and 100 steps per task. CREDENTIALS — never in the goal text. Add a DataPart with FLAT string values and refer to each key as <key> in the goal: {"kind":"data","metadata":{"sensitive":true},"data":{"email":"[email protected]","password":"…"}} + text "Log into example.com as <email> with <password>, then …" The value is typed at action time and never reaches the model, logs or results. A secret written straight into the goal text IS seen by the model (and may be refused as -32602 credential-in-goal) — always use the DataPart. URLs are not substituted: write them plainly. Text that must be pasted exactly: <verbatim name="pitch">…</verbatim>. Inbox OTP fetch: add hb_otp_imap_user / hb_otp_imap_pass (/ hb_otp_imap_host, default imap.gmail.com) to the same DataPart. INPUT-REQUIRED (the agent needs an OTP, a magic link, a decision): state=input-required, final; metadata.input_required = {req_id, question, kind, sensitive, timeout_s, deadline}. While it is still paused, answer with message/send: message.referenceTaskIds=[taskId], message.metadata.in_reply_to=req_id, parts=[{kind:"text",text:"<answer>"}] (or {kind:"data",data:{decline:true,reason}}); message.taskId=<taskId> or re-sending on the same contextId also works. The pause message carries a ready resume_example. A human can answer in the viewer instead. No answer by the deadline = auto-decline. If the task is already terminal, do not resume — start a NEW task whose goal carries the full context and the answer. FILES IN: attach {"kind":"file","file":{"name":"a.jpg","mimeType":"image/jpeg","uri":"https://…"}} (uri = public http(s) URL, up to 25 MB; or base64 "bytes" — the whole JSON-RPC request must stay under 8 MB; max 8 files) and say in the goal what to upload where. OPTIONAL metadata (params.metadata or message.metadata): country (iso2, e.g. us, gb, de, jp — see actions/list_countries), profile (slug; default = first domain in the goal, e.g. adsy.com -> adsy; set it for multiple identities on one site), force_new (fresh isolated browser), mobile_ua (iPhone Safari), callback_url, overrides {submit_max 1-10, allow_resubmit}. Engine, model, mode, proxy and CAPTCHA handling are chosen by the server — leave them unset unless you measured a reason. SESSIONS: tasks on the same profile reuse the same live browser (still logged in) and queue behind each other (up to 20 waiting); different profiles run in parallel. Up to 10 concurrent sessions per token. ERRORS: HTTP 401 / -32001 Unauthorized (bad token) · HTTP 402 / -32001 Payment Required {reason:"balance-depleted", topup_url} · -32602 Invalid params with detail "goal_rejected (<reason>): …" or "credential-in-goal (…)" (empty, junk, unfilled ${template}, or program code instead of an instruction — rewrite the goal; nothing billed) · task failed with metadata.error.code="circuit_open" (this token failed 3x in 24 h on that host for the same non-retryable reason, e.g. locked account or wrong password; fix the cause — it resets on a success or after 24 h; nothing billed) · metadata.error.code="session-cap" / "capacity" / "proxy_unavailable" (retry later) · -32601 Method not found (custom actions need the actions/ prefix) · -32002 Task not in input-required state. On failed, read metadata.outcome.result, metadata.outcome.blocks ({code, retryable, how_to_retry}) and metadata.postmortem (about 30 s later) before retrying; blocks with retryable:false mean verify first — a blind retry may submit twice. Report only what these fields say; do not guess "blocked by a CAPTCHA". ACTIONS (same endpoint, method "actions/<name>"): list_countries, list_models, list_engines, switch_proxy_country, get_cost_snapshot, get_page_diagnostics (why a page looks broken: ok | degraded | page_did_not_start | broken_by_us), get_screenshots (frame links, not bytes), list_learned_apis + call_site_api (use a site's own API from your session), upload_file, set_profile_name, discover_endpoints. See customActions. MCP: the same agent at https://agent.humanbrowser.cloud/mcp (Streamable HTTP, Bearer token or OAuth 2.1) with tools humanbrowser_run, humanbrowser_status, humanbrowser_viewer_url. Full reference: https://humanbrowser.cloud/llms-full.txt

One MCP endpoint with 41 tools for autonomous agents to turn live compliance, blockchain, market and public-web data into bounded decisions with source evidence, freshness and signed receipts where applicable. Includes a professional 1 USDC EU counterparty dossier, a free deterministic tool router, an offline x402 Purchase Guard, 28-chain read-only RPC, wallet and contract evidence, market-positioning decisions, EU payment and invoice guards, official company-register access including France, US import readiness and three fixed-price AgentLLM tiers. Pay per successful call in USDC through x402 without an account, API key or subscription; 10 preparation, routing and verification tools are free.

External application-level evidence gate for agent workflows. Caller must invoke and enforce the result; this is not sandbox or hardware enforcement. Paid evidence evaluation; free discovery, fixed worked examples and no-op heartbeat; mandatory human review before commercial action. No independent factual verification or clearance. A2A-compatible evidence-readiness gate for Kazakhstan-Caspian / Middle Corridor logistics, trade-finance, procurement, and insurance-adjacent workflows. Bring route, cargo, counterparties, and dated sources; get structured deal-risk triage, missing source categories, evidence gaps, watch-next indicators, decision-readiness score, risk signal, human-review routing, and named sanctions-exposure flags: OFAC named-sector (EO 14024/14114) and newly-formed-counterparty red flags, EU re-export / circumvention-watch jurisdictions (incl. the 20th-package measures on Kyrgyzstan), dual-use cargo screened against the EU/US Common High Priority List, a domestic-legal vs foreign-sanctions exposure decomposition, and a vessel deceptive-shipping-practice checklist. Deterministic rule-based logic, no model in the decision path. Presence-flagging and evidence triage only, not a sanctions determination. New to these gates? Start with the Corridor & Sanctions Risk Assistant (https://corridor-sanctions-assistant-a2a.vassiliy-lakhonin.workers.dev) for plain-language orientation. Describe the route or counterparty and the next decision or review; fit, scope, fee, and timing are confirmed before work starts.

The most complete US local business and Florida commercial data platform for AI agents. 6.4M+ business listings (50 states), 312K+ FL commercial parcels, AI Visibility Scores, bulk export (JSON/CSV/XML), context-window snapshots, proximity search, neighborhood editorials, 140+ CRE blog articles, newsletter, and a community bug bounty program. 23 MCP tools. x402 micropayments live — 100 free queries/day, then $0.001 USDC on Base.