Skip to main content
Glama
This connector has been deprecated

This connector has been replaced by xrpl-referee.

Create Escrow Vault

create_escrow_vault
Destructive

Create an XRPL escrow vault. Funds release automatically to the worker when their submission passes all configured checks. Free to try: pass network='testnet' (see create_testnet_wallet / try_testnet_demo).

Buyers can require NFT ownership, an atomic NFT Delivery-vs-Payment swap, domain verification, or W3C Verifiable Credentials before a PASS releases escrow — set the relevant proof-gate params below.

Three release modes:

  • AI audit (default): worker submits text/files; AI referee scores against task_description or criteria.

  • Proof-gate only (require_ai_audit=False): payment releases when all configured proof gates pass. No AI call.

  • Delayed release (release_mode='challenge_window_hrs:N'): PASS holds for N hours then auto-releases. During the window, the buyer can challenge via prepare_challenge_escrow + challenge_escrow_verdict. Challenge fee ~$1 USD in XRP locked to the seller; paid to seller if PASS upheld, refundable if overturned or inconclusive.

Structured criteria (recommended for complex work):

  • Pass criteria as a list of strings or typed dicts ({type, description, [gate]}).

  • Types: 'judgment' (AI decides), 'evidence' (verify URL content), 'code' (deterministic gate; supported gates: sha256_of, exact_match, regex, json_schema).

  • INCONCLUSIVE verdicts do not consume an attempt for the first 3 per vault; subsequent INCONCLUSIVEs count as an attempt.

White-label / headless integration:

  • Pass callback_url to receive webhook POSTs on all state changes.

  • Pass metadata (JSON string) to attach your own reference IDs.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "escrow_id": "<escrow_id>",
      "task_description": "<task_description>",
      "buyer_name": "<buyer_name>",
      "buyer_address": "<buyer_address>",
      "worker_address": "<worker_address>"
    },
    "name": "create_escrow_vault"
  }
}

Input Schema

NameRequiredDescriptionDefault
networkNo'mainnet' (default) or 'testnet'. On testnet no fees are charged, vault IDs are prefixed TN-, and explorer links point to testnet.xrpl.org.mainnet
nft_dvpNoAtomic NFT swap: worker must transfer a specific NFT to the buyer before payment releases. On PASS the vault enters PASS_AWAITING_NFT; worker then registers their NFTokenCreateOffer via POST /escrow/{id}/nft-offer and payment auto-releases when buyer accepts. Mutually exclusive with require_nft_proof.
categoryNoMarketplace category for this job. One of: default, creative, code, data, data_analysis, bug_bounty, legal, supply_chain.default
criteriaNoOptional list of evaluation criteria for the AI referee. Each item is either a plain string (treated as a judgment criterion) or a dict with fields: id (optional), type ('judgment'|'evidence'|'code'), description (required), and for 'code' type: gate ('sha256_of'|'exact_match'|'regex'|'json_schema') plus gate-specific fields. (http_status and xrpl_tx gates are coming soon.) Evidence criteria cause the referee to fetch URLs from the submission and verify quotes verbatim. Code criteria run deterministic gate checks before the AI is called. INCONCLUSIVE verdicts do not consume an attempt for the first 3 per vault; subsequent INCONCLUSIVEs count as an attempt. Example: [{"type":"judgment","description":"Code compiles"},{"type":"evidence","description":"PR link shows merged status"}]
currencyNoCurrency to lock. Use "XRP" (no trustline needed) or "RLUSD" (USD-pegged stablecoin).XRP
fee_hashNo64-character hex transaction hash of the $0.10 payment (XRP/RLUSD) to rmcSrkpZ2i2kuvtCPeTVetee9SixP4djR. Omit to use free tier if eligible (new agent wallets and new Pro accounts get 3 free escrows, each including the AI check).
metadataNoOptional dict of key/value pairs to attach to this escrow and echo back in every webhook payload. Use to round-trip your own reference numbers (invoice_id, po_number, tenant_id, order_ref, etc.) without storing them on AgentTrust's side. Example: {"invoice_id": "INV-2026-0042", "po_number": "PO-9981"}.
escrow_idYesUnique receipt code for this vault, e.g. AT-7X9K-2MQ4. Used to reference the vault in subsequent calls.
amount_xrpNoAmount of XRP to lock in escrow. Required when currency is XRP. Minimum: 0.000001 XRP (1 drop — XRPL EscrowCreate minimum). Practically, ensure the bounty exceeds the $0.10 protocol fee.
buyer_nameYesName or identifier of the buyer posting the job.
amount_rlusdNoAmount of RLUSD to lock in escrow. Required when currency is RLUSD.
callback_urlNoHTTPS endpoint on your server to receive webhook POST notifications on escrow state changes (funded, PASS, FAIL, expired). Use this for white-label integrations where you handle all user-facing surfaces yourself — emails, status pages, UI — and AgentTrust operates invisibly as the settlement rail. Payload: {escrow_id, event, verdict, timestamp, metadata}.
external_refNoReference to work found on another platform, e.g. 'upwork:123', 'github:org/repo#45', 'moltbook:post/abc'. Stored on the vault and returned on GET /escrow/{id}. No parsing — stored as-is, max 500 chars.
external_urlNoURL of the original job listing or work source (must start with https://). Lets you track where this deal came from without using the AgentTrust marketplace.
proof_policyNoWhen multiple proof gates are set: 'ALL' (default) requires every gate to pass; 'ANY' requires at least one gate to pass.ALL
release_modeNo'auto' (default): payment releases immediately on PASS. 'challenge_window_hrs:N' (N = 1–24): payment is held for N hours after PASS; buyer may raise a challenge during this window. Payment auto-releases at deadline if no challenge.auto
buyer_addressYesXRPL wallet address (r...) of the buyer.
project_labelNoOptional human-readable label for the job, shown in the marketplace.
worker_addressYesXRPL wallet address (r...) of the worker who will receive payment on approval. Use the address returned by award_job().
max_submissionsNoNumber of work submission attempts the worker is allowed before the vault is locked. Default 3 (included in the $0.10 creation fee). Each slot above 3 costs an extra $0.05 at creation time (e.g. max_submissions=5 → $0.10 + 2×$0.05 = $0.20). Range 1–10.
required_domainNoWorker must have their XRPL wallet domain field pointing to this domain (verified via xrp-ledger.toml). Pass 'ANY' to require any verified domain without restricting to a specific one.
cancel_after_hrsNoHours until the buyer can reclaim funds if the worker does not deliver. Default 168 = 7 days.
deliverable_typeNoType of deliverable: 'work' (default, general text/files), 'code' (software), 'data' (dataset/analysis), 'nft' (NFT token).work
require_ai_auditNoDefault True. Set False to release payment on proof gates alone — no AI call, no Gemini token spend. Requires at least one proof gate (require_nft_proof, required_nft_issuer, required_domain, or required_vc_issuer_did). Use for machine-verifiable deliverables: NFT delivery, domain verification, W3C credentials.
required_vc_typeNoIf set alongside required_vc_issuer_did, the VC must also have this credential type (e.g. 'CertifiedDeveloper'). Leave blank to accept any credential type from the issuer.
task_descriptionYesDetailed specification the worker must fulfil to be paid. Be precise — the AI referee evaluates against this.
counterparty_typeNoWhether the worker is an agent or a human. 'agent' (default) or 'human'. Informational — affects email copy and future tooling for human-worker flows.agent
require_consensusNoRequire consensus between Gemini Flash AND Gemini Pro before a PASS is issued. Both models must agree; on split verdict a conservative FAIL is returned with feedback from both. Fee: $0.25 (vs $0.10 standard). Ignored when require_ai_audit=False.
require_nft_proofNoWorker must own an NFT from any verified issuer (or from required_nft_issuer if set) to receive payment. Set required_nft_issuer to restrict to a specific issuer wallet.
required_nft_issuerNoRestrict NFT proof to NFTs minted by this XRPL wallet address. Also implicitly sets require_nft_proof=True. Leave blank to accept NFTs from any trusted issuer.
required_vc_issuer_didNoWorker must present a W3C Verifiable Credential JWT issued by this DID (e.g. did:web:issuer.example.com). Used for accreditation, certifications, or KYB checks.
require_consensus_vaultNoRequire both Gemini Flash AND Gemini Pro to agree on every evaluation for this vault. One-time vault creation fee of $0.25. Both models must return PASS or the result is INCONCLUSIVE.

Output Schema

NameRequiredDescriptionDefault

No arguments

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.9/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description provides rich behavioral context that goes significantly beyond the structured annotations. It details the financial logic of the $0.10 protocol fee, the attempt consumption rules for 'inconclusive' verdicts, the mechanics of the challenge window, and the specific structure of the webhook payload for white-label integration.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is extremely verbose due to the inherent complexity of the tool (32 parameters, multiple operational modes). However, it is highly effective and well-structured using clear headings (e.g., 'Three release modes', 'Structured criteria') to guide the AI through the large volume of information without becoming a monolithic block of text.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity, the description is highly complete. It covers all technical bases—network differences, multiple payment currencies, diverse proof-gate mechanisms (NFT, VC, Code), and external integration methods—providing a single, comprehensive reference for the tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

While the schema is 100% covered, the description adds critical semantic meaning. For instance, it explains the practical consequences of setting `max_submissions` (extra costs), the intent behind `require_ai_audit=False`, and the specific cryptographic requirements for `code` criteria, which helps the AI interpret the data types correctly.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly goes beyond 'creating a vault' to explain its core function: creating an XRPL escrow where funds release automatically upon task completion. It distinguishes itself by detailing the sophisticated mechanisms of fund release, which is central to its function in the broader toolset.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It provides explicit guidance on usage scenarios, contrasting the three release modes (AI audit, Proof-gate only, Delayed release). It also offers practical usage notes, such as the free tier eligibility for `testnet` and the specific prerequisites for white-label integration.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.