{
  "markdown": "<p align=\"center\">\n  <img src=\"https://vaaya.ai/icon.png\" width=\"96\" height=\"96\" alt=\"Vaaya\" />\n</p>\n\n# Vaaya MCP — payments for any agent\n\nMCP server for AI agent payments. An agent in Cursor, Claude Code, Claude\nDesktop, Codex, or any MCP host can call paid APIs (search, scraping, contact\ndata, public records, compute, media, research) from one balance, with a spend\nceiling it states on every call, and no vendor keys anywhere in its\nenvironment. Billed only on success (x402 USDC on Base, Stripe MPP, or Tempo);\na refused or failed call costs nothing.\n\n```bash\nnpx @vaaya/mcp install            # Cursor, Claude Code, Claude Desktop, Codex\nclaude mcp add --transport http vaaya https://vaaya.ai/mcp\n```\n\nFirst paid call: `consult` returns a priced plan; `use({ service, action,\nparams, max_cost_cents })` runs one line of it. If the quoted price is above\n`max_cost_cents`, the call is refused before it reaches the provider and the\nreason comes back:\n\n```json\n{ \"error\": \"over_max\", \"message\": \"price is 5¢ but max_cost_cents was 1\", \"price_cents\": 5 }\n```\n\nRunnable examples: [`examples/pay-for-an-api`](examples/pay-for-an-api) (one\ncall allowed, one refused), [`examples/cursor`](examples/cursor),\n[`examples/claude`](examples/claude). Guides:\n[give an agent a budget](https://vaaya.ai/blog/give-an-ai-agent-a-budget),\n[Cursor setup](https://vaaya.ai/blog/cursor-agent-mcp-payments),\n[Claude setup](https://vaaya.ai/blog/claude-mcp-payments),\n[what is an AI agent wallet](https://vaaya.ai/blog/what-is-an-ai-agent-wallet).\n\n## What your agent can do through Vaaya\n\n| Capability | What you get |\n| --- | --- |\n| 🎨 **Media generation** | Images & video with top models — Nano Banana Pro 2, GPT Image 2, Seedream v4.5, Kling v3, Seedance 2.0 — plus TTS voices & music |\n| 🎬 **Product demo videos** | Record a raw, silent screen capture of your live product; Vaaya auto-authors narration + zoom/cut plan and renders a finished narrated demo |\n| 🖼️ **Website imagery** | Derive a site's brand and generate an on-brand image set straight into your repo |\n| 🔎 **Web search** | Exa & Parallel: query search, URL contents, async deep research tasks |\n| 🕷️ **Web scraping** | Firecrawl scrape / crawl / map / search / extract, persisted to files |\n| 📚 **Deep research** | Multi-hop, source-verified, cited research reports |\n| 📊 **Market & competitive research** | Company reports (people, hiring, ads, reviews, SEO/LLM visibility), product & feature catalogs with review sentiment, UX research with interactive product maps, \"best X for my data\" bake-offs, persistent competitive knowledge repos |\n| 🎯 **GTM & sales** | Find ICP-matched leads with verified contacts, enrich contacts, track buying signals (funding/hiring/launch/press), 24×7 LinkedIn discovery, segment-based outbound email — every draft held for human approval |\n| ⏰ **Monitoring workers** | Scheduled watches (30 min–weekly) over competitor pricing, deals, reviews, job postings, news — only new/changed findings surface |\n| 🖥️ **Compute sandboxes** | Per-second-billed code execution (Modal) for benchmarks, untrusted code, GPU inference, data jobs |\n| 🌐 **Browser automation** | Browserbase sessions to click, type, log in, fill forms |\n| 📧 **Email** | AgentMail inboxes your agent owns — send and receive replies |\n| 🧠 **Memory** | Persistent cross-session memory via mem0, Zep, or Letta |\n\n## Install\n\n### One command (recommended)\n\n```bash\nnpx @vaaya/mcp install\n```\n\n(`npx vaaya-cli install` runs the same installer.)\n\nSets up the Vaaya MCP server for every agent it detects on your machine —\nClaude Code, Claude Desktop, Cursor, Codex — plus the agent skill.\nIdempotent: re-run any time to update. Then restart your agent; the first\nVaaya call opens a short browser approval (sign-up happens right there).\n\n### Remote (Streamable HTTP + OAuth)\n\n<a href=\"cursor://anysphere.cursor-deeplink/mcp/install?name=vaaya&config=eyJ1cmwiOiJodHRwczovL3ZhYXlhLmFpL21jcCJ9\"><img src=\"https://cursor.com/deeplink/mcp-install-dark.svg\" alt=\"Add Vaaya to Cursor\" height=\"32\" /></a>\n\n```json\n{ \"mcpServers\": { \"vaaya\": { \"url\": \"https://vaaya.ai/mcp\" } } }\n```\n\nClaude Code:\n\n```bash\nclaude mcp add --transport http vaaya https://vaaya.ai/mcp\n```\n\nOpenClaw:\n\n```bash\nopenclaw mcp add vaaya --url https://vaaya.ai/mcp --transport streamable-http --auth oauth\nopenclaw mcp login vaaya\n```\n\nHermes (`~/.hermes/config.yaml`):\n\n```yaml\nmcp_servers:\n  vaaya:\n    url: \"https://vaaya.ai/mcp\"\n    auth: oauth\n```\n\nOn first use your client runs the OAuth flow in the browser; approve and you're\nconnected. Revoke anytime at [vaaya.ai/connected-apps](https://vaaya.ai/connected-apps).\n\n### npm stdio shim (for stdio-only clients: opencode, …)\n\n```json\n{ \"mcpServers\": { \"vaaya\": { \"command\": \"npx\", \"args\": [\"-y\", \"@vaaya/mcp\"] } } }\n```\n\nThe shim ([`@vaaya/mcp`](https://www.npmjs.com/package/@vaaya/mcp)) opens the\nOAuth flow on first call, stores a refresh token locally (`0o600`), and proxies\nthe live tool list from the backend — new server-side tools appear without a\nshim upgrade. It also ships an agent skill so your client routes every\ncapability gap through `consult` automatically.\n\n### Agent skill (works across 70+ agents)\n\n```bash\nnpx skills add vaaya-ai/vaaya-mcp\n```\n\nInstalls the [`vaaya` skill](skills/vaaya/SKILL.md) for Claude Code, Codex,\nCursor, Gemini CLI, Copilot, and any other client that supports the open\n[Agent Skills](https://agentskills.io) standard. The skill bootstraps the\nVaaya MCP server if it's missing and keeps the install updated.\n\n### As a plugin\n\n- **Claude Code**: `/plugin marketplace add vaaya-ai/vaaya-mcp` then `/plugin install vaaya@vaaya` — bundles the remote MCP server, the `vaaya` skill, and the SessionStart reminder hook\n- **Codex**: `codex plugin marketplace add vaaya-ai/vaaya-mcp` then install `vaaya` from the Plugins panel\n- **Gemini CLI**: `gemini extensions install https://github.com/vaaya-ai/vaaya-mcp`\n\n## How it works\n\n1. **`consult(intent)`** — describe any goal in plain English (\"generate a hero\n   video\", \"find 50 heads of RevOps with verified emails\", \"watch my\n   competitor's pricing page\"). It converses, clarifies, and returns the exact\n   call(s) to run.\n2. **`use(service, action, params, max_cost_cents)`** — execute the call.\n   Billed on success; `max_cost_cents` is a hard spend ceiling. Long jobs\n   return `{ async: true, job_id }` — poll with **`result(job_id)`**.\n3. **`session` / `close`** — per-second-billed code sandbox sessions.\n\nPlus a full GTM suite (`gtm_leads_find`, `gtm_lead_enrich`, `gtm_signal_create`,\n`gtm_segments`, `gtm_message`, `gtm_replies`, …) and scheduled monitoring\nworkers (`worker_create`, `worker_findings`, …) — the tool list is proxied live from the backend.\n\n## Example prompts\n\n- \"Generate a hero video for our landing page in 16:9.\"\n- \"Record my product and turn it into a narrated demo video.\"\n- \"Find 50 heads of RevOps at Series-B SaaS companies with verified emails.\"\n- \"Watch my competitor's pricing page and tell me when it changes.\"\n- \"Do deep research on the agentic payments market with cited sources.\"\n- \"Scrape these 200 product pages into structured JSON.\"\n- \"Run this benchmark in a GPU sandbox.\"\n\n## CLI\n\nThe [`vaaya-cli`](https://www.npmjs.com/package/vaaya-cli) CLI manages everything from\nthe terminal:\n\n```bash\nnpx vaaya-cli install        # set up the MCP server for every detected agent\nnpx vaaya-cli status         # connection state + live tool count\nnpx vaaya-cli consult \"...\"  # ask Vaaya how it would do something (returns the plan)\nnpx vaaya-cli reauthorize    # re-run the browser auth flow\nnpx vaaya-cli logout         # disconnect & delete local credentials\n```\n\n## Security & billing\n\n- Your agent never sees vendor API keys or upstream URLs — Vaaya proxies every\n  call server-side.\n- OAuth 2.1 (PKCE, dynamic client registration); anonymous `tools/list` for\n  discovery; revocable grants.\n- Pay-per-call on x402 (USDC on Base), Stripe MPP/SPT, or Tempo. Failed calls\n  are never charged. Every call carries a `max_cost_cents` guard.\n\n## Links\n\n- Website & docs: https://vaaya.ai\n- Agent-readable docs (llms.txt): https://vaaya.ai/llms.txt · full tool reference: https://vaaya.ai/llms-full.txt\n- npm: https://www.npmjs.com/package/@vaaya/mcp · https://www.npmjs.com/package/vaaya-cli\n- Support: support@vaaya.ai\n",
  "bytes": 8321,
  "sha": "59bfb2106828111823144bbd3974f1526a5d7bd563f5aee688e412f3ea46a540",
  "repo_slug": "vaaya-ai/vaaya-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_vaaya_ai_vaaya_mcp_759f946b/readme"
}