X402 Trust Mcp — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited X402 Trust Mcp (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
An MCP server for discovering x402 endpoints and checking their trust before paying them: semantic and keyword search across the monitored catalog, plus trust scores, history, bulk scoring, and watches for any endpoint.
Backed by x402-trust.com, which monitors the x402 ecosystem on Base and Solana 24/7: uptime probes, 402-envelope spec compliance, price history, and real on-chain USDC settlement volume per endpoint. Every score is computed deterministically, no LLM involved, so the breakdown is fully auditable.
| Tool | Cost | Description |
|---|---|---|
x402_ecosystem_stats | free | Snapshot of the whole x402 ecosystem: endpoint counts, reachability, and real USDC settlement volume over 30 days. |
x402_trust_leaderboard | free | Top-100 ranking of the most trustworthy x402 hosts, scored by their whole measured surface. Page with offset. |
x402_verified_providers | free | Hosts that proved domain control and confirmed a contact email. Informational only, verification never changes the score. |
x402_search | free | Substring search over host, URL and service name across the catalog. |
x402_semantic_search | free | Meaning-based search: describe the capability you need in plain words. Discovery only, no verdicts. |
x402_trust_preview | free | Three complete trust reports (best, median, worst) exactly as x402_trust_score returns them. |
x402_trust_score | paid | The full trust verdict for one endpoint: score 0-100, grade, recommendation, price, and flags. Unknown endpoints get live-probed on the spot. |
x402_endpoint_history | paid | Raw probe and event history for one endpoint over 1-90 days. |
x402_find_alternatives | paid | Semantically similar endpoints that out-score a given one. Empty result means nothing beats it. |
x402_trust_bulk | paid | Score up to 500 endpoints in one call; the cheapest fitting tier (10/50/100/200/500) is picked automatically. |
x402_watch_create | paid | 30-day monitor for one endpoint: alerts on payTo or price changes, delisting, spec regressions and downtime, via polling or webhooks. Returns a one-time secret. |
x402_watch_events | free | Poll a watch's event log with watch id and secret, paging by the returned cursors. |
x402_watch_edit | free | Change a watch's webhook URLs, sensitivity or subscribed events. |
x402_watch_cancel | free | Stop a watch now; its log stays readable until expiry. |
x402_watch_renew | paid | Add another 30 days to an active watch. Same secret. |
Paid tools cost from $0.005 (single lookup) to ~$0.50 (500-endpoint bulk), charged over x402 (USDC on Base). With X402_PRIVATE_KEY set the server auto-pays within your limits; otherwise it returns the price quote for your host to pay. List tools page with limit (max 25) + offset and return a total.
Field-level documentation, verdict semantics, watch cursor rules and webhook signature details live at x402-trust.com/llms.txt; every response is self-describing via its schema at x402-trust.com/schemas.
Add to your MCP client config (e.g. Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"x402-trust": {
"command": "npx",
"args": ["-y", "x402-trust-mcp"]
}
}
}To enable autonomous payment for the paid tools, add a funded Base USDC wallet:
{
"mcpServers": {
"x402-trust": {
"command": "npx",
"args": ["-y", "x402-trust-mcp"],
"env": {
"X402_PRIVATE_KEY": "0xYOUR_BASE_WALLET_KEY",
"X402_MAX_USD": "0.05"
}
}
}
}| Var | Default | Description |
|---|---|---|
X402_TRUST_API_BASE | https://x402-trust.com | API base URL. |
X402_PRIVATE_KEY | _(unset)_ | Base wallet private key. Enables auto-pay for paid tools. A set-but-malformed key logs a warning and leaves auto-pay off. |
X402_MAX_USD | 0.05 | Per-call auto-pay ceiling. 0 disables auto-pay. |
X402_MAX_TOTAL_USD | 1.00 | Cumulative auto-pay cap per process. 0 = unlimited. |
X402_MAX_CALLS | 1000 | Max paid calls per process. 0 = unlimited. |
X402_TIMEOUT_MS | 20000 | Request timeout. |
X402_BASE_RPC_URL | public Base RPCs | Endpoint for the pre-signing USDC balance check (free eth_call; defaults to the public Base mainnet/Sepolia RPCs). |
The free tools work with no configuration at all.
X402_PRIVATE_KEY is a hot wallet: fund it with only what you are willing to spend on trust lookups. The key never leaves your machine; it signs EIP-3009 payment authorizations locally, after policy checks: chain allow-list (Base mainnet by default), canonical USDC contract verification, and the spend and call-count caps above. Before every payment the client re-reads the wallet's USDC balance: a wallet that cannot cover the quote is refused locally, before anything is signed, so an empty wallet never produces failed settlement attempts, and a top-up takes effect on the very next call.
Signed tool results carry a top-level signature block: an Ed25519 signature over the JCS-canonicalized (RFC 8785) response. It proves the content was assembled by x402-trust and not modified afterwards.
One rule matters when verifying: pin the trust anchor, never follow it from the response. signature.publicKeys is a discovery hint; a forged response would point at the attacker's own key document and still verify. Pin the public key itself or fetch https://x402-trust.com/.well-known/x402-trust-keys.json once at bootstrap and cache it. A worked test vector and a 20-line reference verifier live at x402-trust.com/schemas. Watch management responses are unsigned by design: they carry capability secrets that must never be forwarded as evidence.
MIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.