phoenix-execution-check

Before sending an on-chain transaction, swap, bridge or x402 payment on an L2 (Base, Arbitrum, Optimism, zkSync, Scroll, Linea, Blast, Mantle, Mode, Taiko, Polygon zkEVM), check whether the network is healthy enough to execute right now or whether to wait. Also get this agent's own behavioural identity passport (agent / human / bot) from Silicon DNA.

Phoenix Zero measures 12 chains every ~2 seconds against 2-3 independent RPC nodes (only real JSON-RPC answers count) and answers one question: send now, or wait? Base URL: https://rtt.phoenix-ai.work

1. Before you send a transaction or payment

Free check (the verdict is delayed ~60 s; limit 100 calls/day per IP):

bash

curl -s "https://rtt.phoenix-ai.work/api/v1/demo/safe?chain=base"

Response: {"safe": true, "reason": "ok", "chain": "base", "delayed_s": 63.6, "as_of": "...", ...}

• safe: true → the network was healthy a minute ago; proceed.
• safe: false → wait 10-30 s and check again, or execute on another chain.
• reason is one of: ok, elevated_latency, high_latency, sequencer_stall, elevated_revert, high_revert, data_stale.

Chains: base, arbitrum, optimism, zksync, scroll, linea, blast, mantle, mode, taiko, polygon_zkevm, casper.

Live verdict (no delay) — paid, $0.01 USDC per call via the x402 protocol (Base, Polygon or Arbitrum):

bash

curl -s -i "https://rtt.phoenix-ai.work/api/v1/safe?chain=base"

Without payment it returns HTTP 402 with the payment terms (x402 v2 PAYMENT-REQUIRED header and the same accepts list in the body). Only use it if you have an x402-capable wallet/tool; otherwise use the free delayed check above. The paid answer adds p99 latency and revert ratio and is signed (EIP-191, signer 0xdac8adC44f621bC3A56E9DC108d2F76Be3142294).

Current price (free, no verdict inside): curl -s https://rtt.phoenix-ai.work/api/v1/price

2. Get your own agent passport

Silicon DNA classifies the caller (you) from your request timing and headers:

bash

curl -s https://rtt.phoenix-ai.work/api/agent/passport

The timing_dna block is computed from your own request intervals (Shannon entropy, Spearman ρ, coefficient of variation). Until you have made ~20 requests it honestly says "verdict": "insufficient". It cannot classify a third party — it always describes whoever calls it.

Rules

• Never treat a delayed safe: true as a guarantee; it describes the network ~60 s ago.
• Do not call more than needed: one check right before sending is enough.
• Full endpoint list: https://rtt.phoenix-ai.work/llms.txt