Agentic Finance Graph
Server Details
Ledger of AI-agent money on Base: payments, detections, Gate, signed Pass, Agent Books, free checks
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- Repository
- AgenticFinanceGraph/agentic-finance-graph-mcp
- GitHub Stars
- 0
- Server Listing
- Agentic Finance Graph MCP Server
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 29 tools
The set contains six similarly purposed 'check' tools (pay_check, gate_check, payee_facts, payer_check, self_check, fleet_check) and a cluster of per-agent record tools (agent_record, agent_pass, agent_statement, agent_books, agent_pattern, agent_stability, agent_independence) that all return overlapping slices of level, payments and incidents. Descriptions are unusually long and explicitly cross-reference each other ('for X use Y'), which rescues most selection calls, but the boundaries are only clear after reading a paragraph rather than from the name.
Two conventions coexist: noun-scoped queries (agent_pass, agent_record, market_state, ladder_dynamics, payee_facts, payer_set) and verb-first actions (find_agent, list_metrics, verify_sum_log, canary_register, since_last). Both are readable, but the 'agent_' prefix is overloaded — it sometimes means 'per-agent analytics' (agent_pattern) and sometimes 'a signed artifact' (agent_pass, agent_statement) — and check tools use a suffix style that mirrors nothing else.
29 tools exceeds the well-scoped range and pushes past the point where each tool clearly earns a slot; several could be folded (agent_pattern/agent_stability/agent_independence, or self_check/fleet_check as one tool with a list argument). The breadth reflects a genuinely large data product, but the surface is heavier than an agent can hold in working memory without the very long descriptions.
Coverage of the domain is strong: discovery (find_agent, ranked_agents), per-agent evidence, verification (verify_sum_log), market aggregates (market_state, list_metrics, metric_history, since_last), safety monitoring (detections, blast_radius, canaries) and payment gating are all present. Minor gaps: canary_register has no revoke/delete counterpart, and there is no dispute or correction path for an agent's record.
Publisher details
- Operator
- Agentic Finance Graph
- Operator website
- Unknown
- Vendor relationship
- Unknown
- Documentation
- Unknown
- Privacy policy
- Not provided
- Trust center
- Unknown
- Restrictions
- Unknown
Available Tools
29 toolsagent_booksAgent Books: one agent's payments and receipts for a period, signed and logged (free)ARead-onlyIdempotentInspect
Use this when someone needs a record they can hand to an accountant, an auditor, a platform or a counterparty: what one registration paid and received on Base in a closed month (period=YYYY-MM) or since its first event (period=to-date, the default). Returns counted payments (count, USD, what left the agent's control, what went into its own positions, payees, split by token, ten largest payees), what was left out and why (routing hops, own transfers, payments before registration), counted receipts (count, USD, senders, the part from wallets of other registered agents), detections logged in the period and transfers of the registration. The statement is signed with Ed25519, its hash is a leaf of our signed statement log and a daily Base transaction commits to that log; the answer says how to check each step without trusting us. One statement per agent, period and day; the answer's permalink returns that same statement on any later day. A record of observed facts under agent.books.v1: not an audit, not an opinion, says nothing about who owns the agent, about tax, or about safety.
Inputs: agent (required), period.
agent_independenceDoes an agent's rank rest on demand outside its own operator set (free)CRead-onlyIdempotentInspect
For one ranked agent: its independence score (1 if it paid, in 30 days, an address with demand from three or more operator sets; else half the best score of an agent outside its set that paid one of its payees, best path, at most three hops), the path, its L9 counterparty count under the published rule and under the cluster-aware rule, the inflation ratio between them, and the size of its operator set (same bound wallet, owner, minter or first funder, each link capped at 50). Definitions rank.independence.v1, rank.inflation.v1, cluster.ratio.v1.
Inputs: agent (required).
agent_passThe Agent Pass: eight checks on an agent at one moment (free, signed)ARead-onlyIdempotentInspect
Use this when you need a signed snapshot to store or show (valid three hours); it is not a prediction. What the ledger measured about one registration at its last evaluation under agent.pass.v2, valid three hours: published level L7+, no open harm-kind incident, a counted payment in the last 30 days, its newest payment in character, no issuer freeze on a bound wallet, no unlimited approval to a plain key, no harm-kind incident in 90 days, and a month without a payment out of character. Answers holds or not, which checks failed, which could not be taken (listed, never written as 0), when it was checked and when it expires, with the issuer's public key. Every Pass that holds is signed with Ed25519 (since 24 Oct 2026); since 7 Nov 2026 the same Pass is also served as a JWT and as a W3C Verifiable Credential (JWT encoding, issuer did:web:agenticfinancegraph.com) at /api/pass?agent=&format=jwt or &format=vc, for systems that verify offline. A Pass says what was checked and when; it never says safe, verified or trusted, and says nothing about the next three hours.
Inputs: agent (required).
agent_patternIs an agent's newest payment in character (free, advisory)ARead-onlyIdempotentInspect
Use this to ask whether an agent's newest payment is in character; on its own it is not a pre-payment check (pay_check is). The out-of-character score of an agent's newest counted payment against its own earlier payments (behaviour.out_of_character.v3 since 7 Nov 2026: amount out of its distribution, a new payee at a new size, a never-paid plain key, a probe then a large payment, a look-alike address, an approval change, off its rhythm), with the reasons that fired, how many of its last 30 days of payments scored notable or major, and the market's rates. Levels are none / notable / major; a level is a pattern for a human to read, never a verdict, and nothing is blocked by us from it (advisory since 1 Nov 2026). Every figure carries its measurement time and definition id.
Inputs: agent (required).
agent_recordAn agent's recordARead-onlyIdempotentInspect
Use this for one ranked agent's full free record; to look up one or many agents by id, name or wallet, use find_agent. The free record of one ranked agent: rank with evidence, bindings, and its 25 newest payments with counterparties. Unranked ids return not found: an agent never seen paying has no record to sell or show.
Inputs: id, agent. Returns: url, note, agent, as_of, +1 more
agent_stabilityHow long an agent has behaved the same way (free)BRead-onlyIdempotentInspect
The behaviour clock for one ranked agent: run_days (complete UTC days since the last break in its own payment behaviour, by a CUSUM on payments, USD, payees and new-payee share against its own history), the last break's date and feature, and its heartbeat over 30 days (scheduled, irregular, bursty or sparse, with the dispersion index and the gap CV). A break is a change, not a harm; the measured false-break rate is published in the document. Definitions stability.run_length.v1 and stability.heartbeat.v1.
Inputs: agent (required).
agent_statementAn agent's latest statementARead-onlyIdempotentInspect
The latest 3-hour statement for one ranked agent: payments out split into spend, routing hops and payments whose receipt is not read yet; what left its control; income; open flags; the hash of the previous statement; and a Merkle proof against the window's Ed25519-signed root. Use this when you need the 3-hour statement with its Merkle proof; for the quick pre-payment view use pay_check, which includes it. A client that checks the statement itself: refuse if stale, if chain_ok is false, or if flags is not empty; then call find_agent and refuse if open_incidents is above zero. A statement covers a closed 3-hour window and publishes a few hours after it closes, so it cannot show the last few hours.
Inputs: id, agent.
blast_radiusWhat a stolen key would control: an agent's wallets at one block (free)ARead-onlyIdempotentInspect
For an agent (L8 or above) or one of its bound wallets: priced stablecoin and token balances and ETH in USD at one block, the EIP-7702 delegate if any with its code hash and known sweeper family, and the open ERC-20 allowances the wallet granted, confirmed on chain and capped at the balance per token (agent.blast_radius.v1). Read every six hours from free public nodes. A measurement of exposure at one block, not a prediction; unpriced tokens are listed by symbol.
Inputs: agent, wallet.
canary_registerRegister a canary (operator session required)AIdempotentInspect
Needs a Desk session (Authorization: Bearer); without one this returns 401. Register the public part of a decoy you planted: kind address|key|url|apikey|tool|x402 with exactly one target (address, url_sha256, key_sha256, tool_name or resource_path), a label, the triggers to watch (nonce, code, permit, siwe, inbound_transfer, http_hit, api_use, tool_call, x402_request) and planted_at. Needs Authorization: Bearer ; the session's wallet is the operator. Private material is refused before validation; a bait key must be fresh (nonce 0, no code). A canary is a decoy you plant where only an intruder or a misled agent would read it: a bait key that never holds a balance (its first use is the signal), an address planted as a treasury, a URL, an API key, a tool description or a paid x402 resource. Register the PUBLIC part only (address, hash, tool name, path); we refuse anything that looks like a key, a URL or an API key in clear. We read the chain every 2 minutes (nonce, code, gas funding, inbound transfers, permits) and your own channels when you push them (canary ingest). A fire is a pattern for a human to read, written to your canary log; nothing is published about a plant. Live since 1 Nov 2026 (shadow 16 Oct - 1 Nov): at most 25 live canaries per operator; fires reach you through canary_status, not by e-mail yet.
Inputs: registration (required).
canary_statusYour canaries, their state and their fires (operator session required)ARead-onlyIdempotentInspect
Needs a Desk session (Authorization: Bearer); without one this returns 401. Every canary you registered, with the watcher's last reading (nonce, code, balance, checks) and the newest 20 events per canary (trigger, severity, evidence, whether you declared it a test, the reading in plain words and what to do). Needs Authorization: Bearer . Labels are your own text and come back in the untrusted envelope.
Inputs: no named parameters.
counterparty_previewWho paid this address (free preview)ARead-onlyIdempotentInspect
For one receiving address on Base (or an x402 resource URL listed in the Bazaar): how many counted payments registered agents made into it, how many distinct agents and how many at L7 or above, first and last seen, and the routing hops excluded. Counts only; the payer list with levels, binding confidence and transaction hashes is the counterparty pack. An address that only received routing hops is not a payee. Without an address: the endpoints the most ranked agents pay.
Inputs: address.
definitionMetric definitionARead-onlyIdempotentInspect
The frozen definition of a metric: what it counts, the method, what it excludes and what it does not claim. Use the definition ids returned by the other tools.
Inputs: id (required).
detectionsDetectionsARead-onlyIdempotentInspect
Anomalies our detectors logged, newest first: drain-shaped payments (each with the share of the wallet's balance it took), retry storms, mint bursts, fleet-funding fan-outs (layering_fan), owner changes under a binding, delegation to unknown code. A detection is a shape with evidence, never a verdict. Filter by agent, kind, severity, time or live incidents only; compact drops the glossaries.
Inputs: kind, agent, limit, since, +3 more.
evidence_pack_quoteQuote an evidence packARead-onlyIdempotentInspect
Price and how to buy a dated evidence pack over x402 (USDC on Base, 1 to 30 cents). About one period: edition (the depth file behind a final State of Agent Money edition: every agent and every payee of the period, the edition's figures and the period's book; id = YYYY-MM, YYYY-Qn or latest; 30 cents). About one agent: precheck (the routine check before you pay or hire it: level, last payment, live incidents, statement freshness, and with payto what that address is), liveness, paygraph (where it pays), drain, incidents, peer, builders (which apps build its transactions, ERC-8021), exit (did its wallet empty out: every window in which a new payee took most of its outflow in 72 hours, with where the money went next; candidates, never verdicts), trace (where did the money go: its exits followed live up to three hops, with each hop's issuer freeze, swap out of a freezable token and Tron activity), bundle, reval (a live re-probe) or statement (its last 24 statements with proofs). About one address: counterparty (which ranked agents paid it) or venue (what the address is: label, contract or key, who pays it, poisoning look-alike or not, live 48-hour flow). This tool does not pay; an x402-capable client fetches the returned URL, receives HTTP 402 with the exact price, and retries with a signed payment.
Inputs: kind (required), id, agent, payto.
find_agentFind an agentARead-onlyIdempotentInspect
Use this to look up one or many agents by id, name or wallet; agent_record gives one ranked agent's full record. Look up an ERC-8004 agent by token id on Base, slug, wallet address (0x...) or the name it declares. Returns whether it is ranked, its liveness level (L0-L9), bound wallet, stablecoin float, payment totals and open incidents. Up to 25 ids at once, separated by commas. A name is matched among ranked agents only, and a declared name is the agent's own claim, not verified.
Inputs: query (required).
fleet_checkFleet check: the self-check for many agents at once (free)ARead-onlyIdempotentInspect
Use this when someone runs or depends on several agents: give agents (a list of up to 60 token ids, slugs or bound wallets) or owner (a wallet: the Base registrations it holds now, the 60 newest, with the total). One row per agent with its level, last payment, whether its Agent Pass holds, the Watch reading of its newest payment, open incidents, standing unlimited approvals and approvals to plain accounts, detections in 30 days, blast radius, and the items worth a look; rows with something worth a look come first. Each row is that agent's self_check in short: open it for sources and measurement times. A clean row is not a statement that the agent is safe.
Inputs: owner, agents.
gate_checkCan this agent pay this address, now, this much? (free, advisory)ARead-onlyIdempotentInspect
Use this when your client applies a written payment policy (AP2 vocabulary) and wants block, warn, log or allow with every reason; it is not a safety verdict, and we never block. One call before an agent pays: the payee facts, the agent's own context from our ledger (what it paid in the period, how often, whether and when it first paid this payee, how many new payees today), the out-of-character score of this payment as the agent's next, and a decision under a policy written in the AP2 mandate vocabulary (allowed_payees, amount_range, budget {max, period: daily|rolling_24h|rolling_7d}, agent_recurrence, execution_date) plus ours: new_payee {max, window_hours, max_new_per_day, step_up_ratio, step_up_days}, refuse_conditions {fact: block|warn|log|ignore} and signal {v2: {notable, major}}. Anything the evaluator does not understand fails closed. Named policies: log-only (default), balanced, recommended, recommended-v3, strict (their measured cost on 500 real agents is in the policy document). The answer is block / warn / log / allow with every reason, and is recorded (advisory since 1 Nov 2026; the default policy log-only never warns or blocks): we never block anything; your client decides, and decides its own fail mode when we are unreachable. Since 13 Nov 2026 every answer carries a receipt (gate.receipt.v1): the agent, the payee, the amount, the policy name and hash, the decision, the reasons and the time, signed with Ed25519, to keep beside the payment as proof that it was checked first; receipts of checks made with a Desk session are also written into our signed log, which a daily Base transaction commits to (read one back at /api/gate/receipt?id=<decision_id>). A receipt never says a payment was made or that the payee is safe.
Inputs: agent (required), payee (required), amount_usd (required), policy, +1 more.
ladder_dynamicsHow ranked agents end, and which platforms drain more than the market (free)BRead-onlyIdempotentInspect
The whole-cohort document: cumulative incidence of the four competing exits (drained, abandoned, dormant, none) at 30, 60, 90 and 180 days with the numbers at risk and the restricted mean time without an exit (ladder.exit_incidence.v1); the Standardised Drain Ratio per minting address with 30 or more ranked agents, with 95% intervals (platform.sdr.v1); the day-to-day transition matrix between L7, L8, L9 and out (ladder.transition.v1); and the landmark re-check of the published survival figures (survival.landmark.v1). compact: the figures only, without the rules, caveats and notes.
Inputs: compact.
list_metricsList the measured metricsARead-onlyIdempotentInspect
Every metric the ledger measures or archives: key, unit, definition id, whether it is a headline figure, its first stored sample, and the date from which its history compares across our own changes (comparable_since). Use a filter (part of a key, a definition id or a word from its note) to get the notes too.
Inputs: filter.
market_stateMachine money: headline figuresBRead-onlyIdempotentInspect
Use this for today's headline figures. The ledger's headline figures for AI-agent money, each with its value, unit, measurement time and definition id. On Base: money agents paid to others after registration (counted), daily spending by the day it was paid (last day, 7-day and 30-day averages), the part that went into positions they still hold, what left their control, the routing hops and pre-registration history excluded, how many agents are ranked (L7+ paid someone, L8 funded, L9 repeat counterparties), agent income, x402 seller inflow, treasuries, concentration, detections and the address-poisoning watch. On other chains: ERC-8004 registrations on seven chains, what registration owners paid on BNB Smart Chain, Ethereum, Arbitrum, OP Mainnet and Polygon, and Solana agent registrations and payments.
Inputs: compact. Returns: as_of, chain, since, source, +4 more
metric_historyMetric historyARead-onlyIdempotentInspect
Use this to chart a figure over time. The history of one measured metric from the ledger's archive (daily points, plus 5-minute points for the last 48 hours). Gaps are real gaps, never filled. An unknown key returns similar keys and the full list of archived keys.
Inputs: key (required), range.
pay_checkCheck an agent before paying it (free)ARead-onlyIdempotentInspect
Start here before your agent pays or hires another agent. One free call before your agent pays or hires another: its latest statement (stale, age, chain_ok, flags), live incidents, its detections in the last 72 hours, the token approvals its wallet has given (unlimited ones, and any to a plain account), its last counted payment, and its payments seen after the statement window closed (from its 25 newest payments, so the hours a statement cannot show are not blind). Returns facts and which of the documented refuse conditions are met; the decision stays with you. The paid precheck pack adds what the address you were given to pay is. Not for the address you were given to pay (payee_facts) or for applying your own written policy (gate_check).
Inputs: agent (required).
payee_factsNine facts about an address before paying it (free)ARead-onlyIdempotentInspect
Use this for the address your agent was given to pay; for the paying agent itself, use pay_check. What the chain and our ledger say about an address an agent is about to pay (payee.facts.v2): its code kind (plain key, contract, EIP-7702 delegated) and whether the delegate matches a known sweeper or the tiny-forwarder shape; whether it forwards most of what it receives within minutes (relay shape); whether its stablecoin outflow goes to swap venues; who funds it and where it exits (7 days of Transfer logs); whether a contract exposes a claim-on-proof selector; an issuer freeze; how many registered agents pay it; and, given the paying agent and the amount, whether it is a never-paid destination at a size the agent never sent and whether it is a look-alike of an address in the agent's own history. Each fact carries value, evidence, method and measurement time; an unavailable channel is said, never guessed. Facts describe an address; they do not judge it.
Inputs: payee (required), agent, amount_usd.
payer_checkWho is paying me: what the ledger holds about a paying wallet (free)ARead-onlyIdempotentInspect
Use this when you sell to agents (an x402 seller, a marketplace, an API) and want to know who is on the other side before you serve or after a payment arrives: give wallet, the 0x address that is paying you. Returns whether a Base ERC-8004 registration is bound to it (any bound wallet), how many, the highest-level one in short (level, last payment, whether its Agent Pass holds and until when, the Watch reading of its newest payment, open incidents, detections in 30 days), what the address itself is in our feed (plain key, contract or EIP-7702 delegate, frozen by an issuer, planted as a poisoning look-alike), and the items worth a look, each with its source. A wallet with no registration is reported as such: that is not a statement about it. Facts, never a verdict; you decide whom to serve.
Inputs: wallet (required).
payer_setWho paid an agent, and whether its reviews come from payers (free)ARead-onlyIdempotentInspect
The wallets observed paying one agent (income into its bound wallets, plus counted payments from other registered agents), up to 250, with payments, USD and first and last paid; and how much of its ERC-8004 feedback is payment-grounded (the reviewer paid the agent before rating it). Pass the addresses as clientAddresses to the Reputation Registry's getSummary to read only reviews from wallets that paid. Observed payers are a lower bound, and payment-grounded is not independent: a payer under the same operator grounds its own reviews.
Inputs: agent (required), limit.
ranked_agentsRanked agentsARead-onlyIdempotentInspect
The top ranked agents (L7 and above: a bound wallet was seen paying money out of the agent's control after registration). Sorted by level first (L9, then L8, then L7), then by lifetime counted USD; sort_by changes the second key to distinct counterparties or number of payments. A high place means a large counted spender at that level, not a working service: read agent_record before relying on one. Free tier: up to 25.
Inputs: limit, sort_by, min_level. Returns: as_of, agents, returned, total_ranked, +1 more
self_checkSelf-check: everything the ledger holds about one agent (free)ARead-onlyIdempotentInspect
One free call for an agent to check itself, or for the person who runs it: its record and level, the Agent Pass (holds or not, what failed, what could not be checked), the Watch's reading of its newest payment (none / notable / major with reasons), the token approvals its wallet gave in 30 days (unlimited ones with the money behind them, any to a plain account), its detections in 30 days and, at L8 and above, what a stolen key would control at one block. Ends with worth_a_look: the items a person should read, each with its source. Every part carries its measurement time and definition id. Facts, never a verdict: a clean list does not mean safe.
Inputs: agent (required).
since_lastA figure beside its previous sampleARead-onlyIdempotentInspect
Use this to ask whether a figure moved, and by how much. One measured figure with the stored sample it moved from: current, prior, delta, delta_ratio, both timestamps, definition id, unit, and whether the move is material under the published thresholds. window: sweep (the previous stored sample, the default), 24h, 7d or 30d. No prior sample inside the window returns gap: true and no delta; nothing is interpolated. Accepts a metric key or a definition id.
Inputs: metric_id (required), window.
verify_sum_logVerify our published total and our statement log (skeptical client)ARead-onlyIdempotentInspect
Trust nothing you did not recompute. Verifies the Ed25519 signature on the current sum-log head (the Merkle-sum root over every counted payment and the total it commits to), optionally climbs one agent's proof from its leaf to the root and the whole total, reports how many of the previous head's leaves the current head declares unchanged, changed, removed and added, and verifies the statement log's signed tree head and, given the head you kept (tree_size, root_hash), the consistency proof that our statement history was not rewritten. Keep the heads it returns and pass them next time. Definitions sumlog.v1 and agent.statement.log.v1.
Inputs: agent, previous_head, previous_statement_head.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Agent-native double-entry accounting ledger with x402 micropayments
Agent memory and a signed spending ledger: pay any x402 API, keep a provable receipt.
Proof-of-spend and reserve checks for AI agents: receipts, spend audit, NAV, reconciliation.
Portable client-sealed memory and private rooms for agents from any lab; a free door; a ledger
Related MCP Servers
- AlicenseAqualityBmaintenanceSeals each payment an AI agent makes (x402, Stripe, cards, etc) as a signed record with the company's own key: the agent declares the purchase before paying, checks it against its limits, and records the settlement after. Anyone can verify the records without relying on Luca.10Apache 2.0
- AlicenseAqualityDmaintenanceEnables agents to query a read-only ledger of machine payments, exposing tools for spending totals, purchases with event chains, verdicts, payer passports, and fiscal reporting data.7MIT
- AlicenseAqualityBmaintenancealalanet-agent is the open local client of alalanet.com, the free service an AI agent asks before it installs, connects or pays. A stdio MCP server whose key stays on your machine: it signs agent passports, receipts, skill and package releases, mail and memory locally (16 tools) and checks files and agents against alalanet's public transparency log. Hosted checks: https://alalanet.com/mcp.216Apache 2.0
- AlicenseNot gradedqualityBmaintenanceEnables autonomous financial agents and payment webhooks to create and verify immutable double-entry ledger entries, enforce debit-credit balance invariants, and generate hash-chained audit trails with fraud detection.7MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.