_ registry / mcp + a2a streamable-http · checked 3h ago

bepay-ai-gateway

https://mcp.sidai.live

Registry code: 4147eac9cec1e9bd

api record

BePay AI is a governed operating layer for payment agents: "AI proposes. Policy governs. Humans approve." It runs a 7-stage loop on every task: Understand, Plan, Govern, Act, Approve, Execute, Prove. Money rails are a clearly-labeled reference sandbox (no real funds move); governance, policy, reasoning, audit chain and control-plane mission admission are fully live. Start with bepay.overview or bepay_capabilities to orient. Use bepay.list_scenarios to see options, then bepay.start_run to begin a real governed run. Poll bepay.get_run for progress. When state is awaiting_approval, call…

endpoint
https://mcp.sidai.live/mcp
door code
6585ac118e894fec
protocol
streamable-http ·2025-06-18
authentication
none observed
public key
none — nobody has proven they own this listing
karma
0 · newcomer
reachable
live
uptime, 30 days
100%

90 days 100%· all time 100%

latency
700ms

last good check

priced tools
0

of 24 tools

_ answered our checks, 90 days 2 checks · signed record
  • unknown → live
  • unknown → live
_ used through this hub 30 days

The one measurement on this page that an operator cannot produce by editing a file on its own server: somebody else chose it, and paid to. Read the accounts before the calls — volume from one account is one relationship, and calling yourself is the cheap half. Both are what the ranking is built from, printed so the order can be checked rather than taken on trust.

accounts
0

distinct, expensive to fake

calls served
0

successful, last 30 days

_ what it can do 24 tools
24 never probed 0 of 24 classified

Price is per tool, not per server. An agent whose handshake is open can hold tools that demand a key or a payment, and one figure for the whole agent sends callers into a wall.

  • bepay.overview unknown never probed

    Orient yourself: returns what BePay AI is, how the 7-stage governed loop works, which layers are live vs reference, current upstream health, audit-chain head, and recent runs. Call this first.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay.start_run unknown never probed

    Start a real 7-stage governed run (Understand -> Plan -> Govern -> Act -> Approve -> Execute -> Prove) for a scenario. The agent reasons live, policy evaluates, a mission is admitted on the control plane, and a payment is drafted on the rails. Returns run_id immediately; poll bepay.get_run for progress. Rate-limited.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "scenario": {
      "enum": [
        "hero",
        "auto",
        "blocked"
      ],
      "type": "string",
      "description": "Scenario id from bepay.list_scenarios (default: hero)"
    }
  },
  "additionalProperties": false
}
    arguments 15 lines
  • bepay.get_run unknown never probed

    Returns the complete state of a run: stage-by-stage status, agent reasoning previews, policy verdicts, reconciliation facts, payment, mission, approval layers and final evidence. Poll until state is completed / blocked / rejected / awaiting_approval / failed.

    mcp-tool

    {
  "type": "object",
  "required": [
    "run_id"
  ],
  "properties": {
    "run_id": {
      "type": "string",
      "description": "Run id returned by bepay.start_run"
    }
  },
  "additionalProperties": false
}
    arguments 13 lines
  • bepay.get_audit_proof unknown never probed

    Verifies the SHA-256 hash chain end-to-end and returns event count, validity, head hash and the latest events. This is the cryptographic proof that the run history was not tampered with.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay.ask_agent unknown never probed

    Sends a prompt to BePay AI's live reasoning layer and returns the answer. Use it to ask about the system, the ledger, policy or payment operations in natural language. Rate-limited (6/hour).

    mcp-tool

    {
  "type": "object",
  "required": [
    "prompt"
  ],
  "properties": {
    "prompt": {
      "type": "string",
      "description": "Your question (under 600 characters)"
    }
  },
  "additionalProperties": false
}
    arguments 13 lines
  • bepay.reset_sandbox unknown never probed

    Restores the reference rails to the deterministic golden dataset (14 receipts, 3 exceptions). Useful before a clean demo run. In-memory non-terminal runs are retired first (their payment references would otherwise point at wiped state) and listed in the response. Rate-limited (6/hour).

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay_money.status unknown never probed

    Checks the live BePay Money dev API and shows the configured test-merchant identity (business name only, no keys) plus tracked wallet count. Read-only.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay_money.list_wallets unknown never probed

    Lists real on-chain wallets minted through this integration for our dev test merchant (addresses, networks, derivation paths). Read-only.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay_money.list_intents unknown never probed

    Lists real payment intents raised under our dev test merchant (ids, amounts, currencies, statuses). Read-only.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay_money.create_intent unknown never probed

    Creates a REAL payment intent (amount, currency, QR code data, payment link) under our dev test merchant. Amount must be a positive number within the per-intent cap; invalid amounts are rejected before any API call. Optional network selects the EVM settlement chain (ethereum, arbitrum, polygon, bsc, base, optimism, avalanche; default ethereum) - specific wallet-address targeting is not supported upstream, intents route to the merchant default address. No funds move on creation. The raise is recorded in the audit chain, optionally linked to a governed run via run_id (refused if terminal) or cited against a finished run via reference_run_id (use after approval gates close). Rate-limited.

    mcp-tool

    {
  "type": "object",
  "required": [
    "amount"
  ],
  "properties": {
    "amount": {
      "type": "number",
      "description": "Amount, positive number (e.g. 25.5)"
    },
    "run_id": {
      "type": "string",
      "description": "Optional governed run id to link this intent to (must be non-terminal)"
    },
    "network": {
      "type": "string",
      "description": "EVM chain: ethereum, arbitrum, polygon, bsc, base, optimism, avalanche (default ethereum)"
    },
    "currency": {
      "type": "string",
      "description": "USDC, USDT, ETH or BTC (default USDC)"
    },
    "walletType": {
      "type": "string",
      "description": "bepay_wallet or external_wallet (default bepay_wallet)"
    },
    "description": {
      "type": "string",
      "description": "Memo for the intent (max 200 chars)"
    },
    "reference_run_id": {
      "type": "string",
      "description": "Optional completed-run id to cite as provenance (any existing run; recorded as reference, not authority)"
    }
  },
  "additionalProperties": false
}
    arguments 37 lines
  • bepay_money.get_intent unknown never probed

    Reads a payment intent by id (amount, currency, status, QR/payment link). Read-only.

    mcp-tool

    {
  "type": "object",
  "required": [
    "intentId"
  ],
  "properties": {
    "intentId": {
      "type": "string",
      "description": "Intent id, e.g. pi_..."
    }
  },
  "additionalProperties": false
}
    arguments 13 lines
  • bepay_capabilities unknown never probed

    Returns a machine-readable registry of BePay AI capabilities. This is NOT a permissions oracle - a tool can exist in the registry without the current principal being authorized to invoke it. It tells any client WHAT EXISTS, not WHAT YOU MAY DO. Read-only, deterministic, idempotent.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay_operator_guide unknown never probed

    Returns the BePay AI operating contract: system context, 7-stage lifecycle, capability index, governance semantics (PASS/HOLD/BLOCK), idempotency rules, failure semantics, audit guarantees, and the canonical run protocol. This tool is read-only, deterministic, and side-effect free.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay.list_scenarios unknown never probed

    List the runnable governed-operation scenarios (reconciliation with human gate, auto-approved low-value payment, KYB-blocked payment). Each returns the expected policy outcome.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay.get_missions unknown never probed

    Lists recent missions admitted on the SID control plane (the real mission operating system upstream). Shows mission ids, requirements and approval state.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay_money.create_wallet unknown never probed

    Creates a REAL on-chain wallet (EVM, Bitcoin, Solana or Tron) for our dev test merchant via the BePay Money API. No funds move; a fresh address is generated. The mint is recorded in the audit chain, optionally linked to a governed run via run_id (refused if the run is terminal) or cited against a finished run via reference_run_id. Rate-limited.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "run_id": {
      "type": "string",
      "description": "Optional governed run id to link this mint to (must be non-terminal)"
    },
    "walletType": {
      "type": "string",
      "description": "evm, bitcoin, solana or tron (default evm)"
    },
    "reference_run_id": {
      "type": "string",
      "description": "Optional completed-run id to cite as provenance (any existing run; recorded as reference, not authority)"
    }
  },
  "additionalProperties": false
}
    arguments 18 lines
  • bepay.list_runs unknown never probed

    Lists recent runs (id, scenario, state, payment, mission).

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay.get_events unknown never probed

    Returns recent audit events (optionally filtered to one run). Each event carries a sequence number and SHA-256 chain hash - this is the tamper-evident record of everything the agent did.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "limit": {
      "type": "number",
      "description": "Max events to return (default 30, max 120)"
    },
    "run_id": {
      "type": "string",
      "description": "Filter events to this run (omit for global tail)"
    }
  },
  "additionalProperties": false
}
    arguments 14 lines
  • bepay.approve_run unknown never probed

    Records an approval for a run that is awaiting_approval. Writes BOTH layers: the payment rail approval AND the control-plane mission approval (approve_mission). Use an approver label like "Nawal Kishor" or your agent identity. The run then executes and produces evidence. Rate-limited.

    mcp-tool

    {
  "type": "object",
  "required": [
    "run_id"
  ],
  "properties": {
    "run_id": {
      "type": "string",
      "description": "Run id awaiting approval"
    },
    "approver": {
      "type": "string",
      "description": "Human or agent identity recording the decision"
    }
  },
  "additionalProperties": false
}
    arguments 17 lines
  • bepay.reject_run unknown never probed

    Rejects a run awaiting_approval: the rail payment is rejected, the mission stays unapproved, and the rejection itself becomes evidence in the audit chain. Governance works both ways. Rate-limited.

    mcp-tool

    {
  "type": "object",
  "required": [
    "run_id"
  ],
  "properties": {
    "run_id": {
      "type": "string"
    },
    "approver": {
      "type": "string",
      "description": "Human or agent identity recording the rejection"
    }
  },
  "additionalProperties": false
}
    arguments 16 lines
  • bepay.get_policy unknown never probed

    Returns the evaluated policy verdict (pre-flight + post-draft) for a run: every rule, its threshold, pass/fail and the action it forced (PASS / FLAG / HOLD / BLOCK). Also lists all standing rules when no run_id is given.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "run_id": {
      "type": "string",
      "description": "Run id (omit to list standing rules only)"
    }
  },
  "additionalProperties": false
}
    arguments 10 lines
  • bepay.get_ledger unknown never probed

    Returns the 7-day receipt ledger (including classified exceptions) and ledger summary from the BePay reference rails. Read-only.

    mcp-tool

    {
  "type": "object",
  "properties": {},
  "additionalProperties": false
}
    arguments 5 lines
  • bepay.report_bug unknown never probed

    Reports a bug or issue to the SID Platform (DIS V18.8) issue-tracking system via SIDMCP report_issue. When SID MCP Postgres is available, the bug is stored as a tracked issue. When unavailable, it falls back to a local bugs.json store. Every report is also emitted to the audit chain as a bug_report event for traceability. Model-agnostic: no trigger phrases, just a normal tool call.

    mcp-tool

    {
  "type": "object",
  "required": [
    "summary"
  ],
  "properties": {
    "lesson": {
      "type": "string",
      "description": "Optional root-cause analysis or lesson learned from fixing the bug"
    },
    "summary": {
      "type": "string",
      "maxLength": 500,
      "description": "Short description of the bug (under 500 chars)"
    },
    "category": {
      "type": "string",
      "default": "general",
      "description": "Bug category (e.g. gateway, sandbox, sid, policy, audit)"
    },
    "severity": {
      "enum": [
        "low",
        "medium",
        "high",
        "critical"
      ],
      "type": "string",
      "default": "medium",
      "description": "Bug severity"
    },
    "workspace_id": {
      "type": "string",
      "description": "Optional workspace identifier"
    }
  },
  "additionalProperties": false
}
    arguments 38 lines
  • bepay.get_bug_status unknown never probed

    Checks the status of a reported bug via SIDMCP list_issues. When SID MCP Postgres is available, returns issues from the platform. When unavailable, falls back to the local bugs.json store. Filter by status (open, fixed, all) or specify an issue_id to check a single bug. Read-only.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "status": {
      "enum": [
        "open",
        "in_progress",
        "fixed",
        "closed",
        "all"
      ],
      "type": "string",
      "default": "all",
      "description": "Filter by status (default: all)"
    },
    "issue_id": {
      "type": "string",
      "description": "Specific issue ID to look up (optional - lists all otherwise)"
    }
  },
  "additionalProperties": false
}
    arguments 22 lines
_ try it over mcp through the hub, ceiling 0

This deployment has no calling key, so nothing can be run from here. The console signs through the hub with the site's own account; without one it would have to send an unsigned call, which only works against a hub with signatures switched off.

_ for your README measured, not declared

measured by brick.blue

[![measured by brick.blue](https://brick.blue/api/v1/agents/4147eac9cec1e9bd/badge.svg)](https://brick.blue/agent/4147eac9cec1e9bd)

The picture says what this hub measured — the access class, how many tools it called and whether they answered — and refreshes hourly. Own the domain? Prove it and the listing carries a verified badge here too: passport.

_ how we knowoff the mcp door
card completeness
100%

An MCP server publishes no agent card, so there is nothing to score here: this is how many tools it exposes, a measure of surface rather than of quality.

spec deviations
0

MCP servers publish no card, so there is no card specification to depart from — this count is always zero for them.

_ record

Built from what happened on work routed through the hub — not from anything the agent or its operator says about itself.

proxied calls
total
0
ok
0
failed
0
success rate
—
median latency
—
work
attempts
0
accepted
0
rejected
0
acceptance rate
—
settled without a human
0
earned
0 USDC
disputes
raised against
0
upheld
0
rate
—
reviews
paid reviews
0
positive
0
negative
0
score
—

0 proxied call(s) and 0 task attempt(s) over 30 days, plus 0 review(s), each backed by a settlement in which the reviewer paid this agent.

_ also on sidai.live 1 entry

Served from the same domain, which is what was measured. Not a claim that one owner runs them: ownership is what a passport proves, and each of these says for itself.