# MarketNow

> Trust layer for agent commerce. 8,560 MCP-compatible skills with Sentinel L2 security audits, x402 payments (HTTP 402 + USDC on Base), and AP2 delegated mandates. Human-in-the-loop by default. Maintained by AliceLabs LLC (Wyoming, USA, 2025), founder Edison Flores.

## What MarketNow does

MarketNow is NOT another discovery registry — discovery is solved (MCP Registry, Smithery, Glama). MarketNow solves TRUST: every skill carries a Sentinel security report, declared permissions, and an honest review_status (auto-scanned | human-reviewed | maintainer-verified).

- 8,560 MCP skills indexed across 58 categories
- 43 free skills (no signup required)
- Sentinel L2: real-time security audit (18 Semgrep rules + Docker sandbox + OSV)
- x402 payments: HTTP 402 Payment Required + USDC on Base (chainId 8453)
- AP2 mandates: agents buy autonomously within principal-approved limits
- Human-in-the-loop by default (mandates default to "notify" mode)
- 5 languages: EN, ES, PT, ZH, FR
- Pricing: $0.99–$9.99 one-time, no subscriptions
- License: MNNC-1.0 (source-available, non-commercial)

## For AI agents

MarketNow is designed for autonomous agent consumption. Agents can search, evaluate, and purchase skills programmatically.

### MCP Server
```
npx -y marketnow-mcp
```
Works with Claude Desktop, Cursor, Cline, Continue, Aider.

### Public API (no auth required for reads)
- GET /api/skills-lite.json — Full catalog (8,560 skills, 4.5MB JSON)
- GET /api/search?q=query&category=Data&max_price=2.99&language=es — Server-side search
- GET /api/agent.json — Machine-readable instructions, trust model, schema
- GET /api/agent-ping.json — Live stats and endpoint list
- GET /api/manifest?health — Cache stats, RPC pool health, rate limiter state
- GET /api/free-skills.json — 43 free skills
- GET /api/categories.json — 58 categories with counts
- GET /api/policies.json — Terms, refund, dispute, privacy
- GET /api/openapi.json — OpenAPI 3.1 specification
- POST /api/agent-purchase — Buy a skill (x402 + USDC on Base)
- GET /api/mandates — Delegated mandates (AP2 compatible)
- POST /api/audit-skill — Run Sentinel L1.5 security audit

### Payment
- USDC on Base (chainId 8453)
- Payment wallet: 0x39Dddf5aEdb58A559CF195fB8bdF23F0604Bf5Ee
- USDC contract: 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913
- HTTP 402 Payment Required protocol (x402)

### Trust model
- review_status: auto-scanned (8,517) | human-reviewed (43) | maintainer-verified (0)
- Sentinel L1.5: 6-point MCP security scan (AUTH, injection, validation, CORS, OAuth, rate limiting)
- Sentinel L1.6: 18 Semgrep rules + Gitleaks + OSV-Scanner (live in production)
- Sentinel L2: Docker sandbox with --network none, --read-only, --cap-drop ALL (implemented)
- Risk levels: green (prompt-only) / yellow (network/env) / red (subprocess)

## Links

- [Website](https://marketnow.site)
- [Browse skills](https://marketnow.site/registry)
- [Trust roadmap](https://marketnow.site/trust)
- [Security methodology](https://marketnow.site/security)
- [Open standards](https://marketnow.site/standards)
- [API documentation](https://marketnow.site/handshake)
- [Pricing](https://marketnow.site/pricing)
- [About AliceLabs LLC](https://marketnow.site/about)
- [Terms & policies](https://marketnow.site/policies)
- [GitHub repository](https://github.com/edgarfloresguerra2011-a11y/marketnow)
- [npm package](https://www.npmjs.com/package/marketnow-mcp)
- [Smithery listing](https://smithery.ai/servers/eddyflores100/marketnow)
- [OpenAPI spec](https://marketnow.site/api/openapi.json)
- [Machine-readable agent instructions](https://marketnow.site/api/agent.json)

## Contact

- Email: info@alicelabs.site
- GitHub: edgarfloresguerra2011-a11y
- Company: AliceLabs LLC (Wyoming, USA, 2025)
- Founder: Edison Flores