MarketNow

Trust Layer for Agent Commerce

Discovery is solved (MCP registry, Smithery, Glama). Trust is not.

MarketNow is the trust layer — every skill Sentinel-scanned, every payment verified on-chain, every mandate human-approved by default.

9,248
SKILLS
43
FREE
10
SECURITY CHECKS
5
LANGUAGES

Sentinel L2 Security

Real-time audit on every API call

L1.5 LIVE 6 metadata checks: AUTH, prompt injection, input validation, CORS, OAuth, rate limiting

L1.6 LIVE Code analysis: 18 secret patterns, 18 MCP prompt injection rules, OSV API dependency check, hygiene

L2 LIVE Behavioral: fetches actual source code from GitHub, analyzes 16 runtime patterns (network, filesystem, process, credential, exfiltration, dynamic imports)

All in ONE endpoint, ONE call, real-time. No Docker needed.

How It Works

Agent purchase flow (x402)

$ curl -X POST https://marketnow.site/api/agent-purchase \ -H "Content-Type: application/json" \ -d '{"skillId":"mn-gen-00003","mandateId":"mand_xxx"}' HTTP/2 402 Payment Required www-authenticate: x402 realm="marketnow", chain="base", token="USDC" x-payment-amount: 1990000 x-payment-to: 0x39Dddf5aEdb58A559CF195fB8bdF23F0604Bf5Ee x-payment-token: USDC $ # Agent sends USDC on Base, retries with txHash HTTP/2 200 OK {"mode":"instant_purchase","verified":true,"license":{"key":"MN-GEN-00003-..."}}

AP2 Mandates

Human-in-the-loop by default

✅ Default: notify — human gets email/webhook alert on every purchase

✅ 3 autonomous purchases — agent can buy without per-purchase approval

✅ After 3: requires_reapproval — human must re-approve

✅ Hard caps: $500 total, $50 per purchase, 90-day expiry

✅ Every spend = public git commit at _data/mandates/

Balances agent autonomy with LLM provider safety policies.

10 AliceLabs Security Tools

Original, human-reviewed, $2.99-$9.99

$9.99 PromptArmor Pro
Prompt injection firewall
$9.99 MCP Security Gateway
MCP server firewall + audit
$9.99 RAG Defender
RAG pipeline security
$9.99 AIComply
GDPR/EU AI Act compliance
$7.99 Agentic Red Team
Offensive LLM pentesting
$4.99 AgentMemory Firewall
Memory poisoning defense

Honest Transparency

What we admit we don't have yet

PARTIAL Third-party audit — paid audits deferred until revenue. Volunteer peer review open.

DONE Catalog transparency — 14 bulk-imported categories disclosed, 0 synthetic skills

DONE Human-in-loop by default — not "no humans needed"

PARTIAL Payment reversibility — manual dispute process, on-chain escrow on roadmap Q1 2027

Every status is a git commit. See /trust for the full roadmap.

Try It Now

marketnow.site

$ # Install MCP server for Claude Desktop / Cursor / Cline npx -y marketnow-mcp $ # Browse 9,248 skills https://marketnow.site/registry $ # 43 free skills — no signup https://marketnow.site/registry?filter=free $ # Run a security audit $ curl -X POST https://marketnow.site/api/audit-skill \ -H "Content-Type: application/json" \ -d '{"skillId":"mn-sec-00001"}' $ # GitHub (source-available MNNC-1.0) https://github.com/alicelabs-llc

AliceLabs LLC — Wyoming, USA · 2025
Bootstrapped · No investors · Built in public

1 / 7