AI agent-to-agent payments marketplace reports 100% settlement success on XRPL, Celo; discloses no insurance or audit trail

Agentic Swarm Marketplace, a newly surfaced candidate entry in AI Dispatch's index, sells data and screening services to other AI agents rather than to people. "Agentic Swarm Marketplace sells machine-paid SKUs to other agents — no trading, no speculation," the project's site states. It runs "Three live x402 seller rails for agent-to-agent commerce: XRPL mainnet (T54, XRP), Base mainnet (USDC facilitator), and Celo (native CELO via fulfillQuery )," and each works the same way: "buyers pay per request; the API returns JSON or LLM output after settlement."

On its XRPL rail, the operator says it never takes custody of a buyer's funds: "Receive-only seller ( r... ) — no signing keys on the server." Paid offerings include automated screening of blockchain smart contracts, callable directly by AI tools rather than only by people: "Use the repo’s MCP server so Cursor, Claude Desktop, Smithery, or Glama can call paid SKUs as tools (including contract_triage , contract_audit , airdrop_intelligence ): the bridge pays HTTP 402 invoices via x402_broker_client when your .env is configured."

By the vendor's own account, in self-reported soak-test logs rather than an independent audit, the service recorded "100% Success rate 120 Total XRPL txs 1080 Total Celo txs 68s Avg settlement." The same site cautions against relying on that contract-screening product for high-stakes decisions: "Not legal advice; not a guarantee that a contract is safe to interact with; not a replacement for a dedicated human audit or bug bounty when stakes are high. Always verify independently."

insurance_available, compliance_certs, audit_trail and documented_incidents are recorded as no_public_information in entity_fields for Agentic Swarm Marketplace, evidenced against this 14,103-character capture of agentic-swarm-marketplace.com retrieved Oct. 14 — the capture contains no mention of "insurance" or "indemnification." The entity has not yet been promoted to AI Dispatch's published index and carries no trust-gap scorecard.