OpenChainGraph Suite · ART-02 · Agent Spend-Policy Simulator
v1.0.0

Agent Spend-Policy Simulator

Author a spend policy — merchant allowlists, per-transaction and rolling limits, HNP autonomous-flow rules — then simulate up to 5,000 synthetic agent transactions against it in seconds. Detects scope creep (unauthorized merchants/categories), drip-pattern policy bypass (sequences of sub-limit transactions that cumulatively exceed daily or monthly caps), and unauthorized HNP flows. Paste an ART-01 artifact to anchor into the chain. Exports a full transaction log as CSV.

AP2 Export Chains: ART-01 → ART-02 GPU Parallel Bypass Detection Zero PII · Synthetic Only
🔒 All simulation runs locally. All transactions are fully synthetic. No data leaves your browser. Paste ART-01 artifact JSON only — no real payment data.
Policy presets
Spend policy definition
Merchant allowlist
Leave empty = any merchant. At least one tag → allowlist enforced.
Leave empty = any category.
Spend limits
0 = no daily limit
0 = no monthly limit
HNP (Human-Not-Present) rules
Transactions above this in HNP mode require human review.
Hard ceiling on any single HNP transaction.
Simulation parameters
500
30%
20%
10%

Chaos level injects random violations (scope creep, over-limit, bad payment method, unauthorized HNP). Drip frequency inserts coordinated sub-limit transactions designed to evade the daily cap.

Chain — paste ART-01 artifact (optional)
No parent artifact — chain_depth will be 0. Paste ART-01 output to link chains.
0 / 0
Policy assessment
Policy-bypass paths detected

Drip patterns: sequences of individually-compliant transactions that cumulatively violate the daily limit within the same calendar day. Each path entry shows the sequence length, total cumulative spend, and the transaction IDs involved.

Transaction log (0)
#StatusAmountMerchantCategoryMethodHNPDayViolation

Ask your agent

Copy this paragraph into Claude, OpenClaw, or any MCP-aware agent to run this exact tool, with this sample, and verify the artifact.

Run the AINumbers MCP tool `simulate_spend_policy`. Task: Simulate N synthetic agent transactions against a spend policy definition.
Call it with arguments: {"policy_parameters":{}}
Verify before trusting: this node computes in your browser, so the MCP endpoint returns no execution_hash. Run the tool in the page, export the Policy Mandate artifact it produces, and call `verify_execution_hash` on mcp.ainumbers.co (https://mcp.ainumbers.co/mcp) with that artifact.
Return the ledger link https://ledger.ainumbers.co/ so a human can re-verify without contacting us.
PII rule: All inputs are processed locally in your browser. No data is transmitted. Do not enter real personal data — use synthetic or anonymised inputs only.
Open the tool with the sample prefilled: https://ainumbers.co/chaingraph/art-02-agent-spend-policy-simulator.html#p=v1.H4sIAAAAAAAA_wECAP3_e31Dv6ajAgAAAA