Back to search
100
MCP live MCP 2025-11-25 streamable-http

Second Opinion

com.secondopinionx402/second-opinion

Adversarial verification for AI agents - pay an independent skeptic per verdict in USDC via x402.

Uptime
100.0%
1 direct probes · 30d
Response
1096ms
last probe
Tools
6
callable
Resources
1
readable
Prompts
2
available
Primary URL
https://secondopinionx402.com/mcp

Tools · 6

second_opinion_quick

Independent reasoning-only check of a claim ($0.02, ~20s, no web research). Use for logic, internal consistency, and stable established knowledge (definitions, architectures, history, maths). Returns …

second_opinion_verify

Independent adversarial verification with live web research ($0.20, ~60s). A separate AI attempts to refute the claim using current sources and returns a verdict with citations. Use before irreversibl…

second_opinion_panel

Three independent adversarial verifiers attack the claim through different lenses (evidence, logic, currency) and vote ($0.75, ~90s). Use for high-stakes or contested claims where a single check is no…

second_opinion_try

FREE TRIAL — no wallet, no payment, no account: the same quick reasoning-only check as second_opinion_quick, 3/day per caller (claims up to 500 chars; global daily cap applies). The trial verdict is u…

second_opinion_result

Fetch the verdict for a paid async job — polling is FREE and the verdict is already paid for. Returns status running|done|failed, with the full verdict when done.

second_opinion_pricing

FREE: current tiers, prices, and exact on-chain payment requirements (accepts) for every rail — Algorand, Base, Polygon, Solana. Use this to pre-build a payment, or to discover what a verification wil…

Resources · 1

verification-policy

Paste-ready standing policy for agents: verify the load-bearing claim before irreversible actions; refuted = hard stop, inconclusive = escalate or hold, supported = proceed and keep the signed receipt…

secondopinion://policy

Prompts · 2

adopt_verification_policy

Adds Second Opinion's drop-in verification policy to the conversation as a standing rule: get an independent adversarial check of load-bearing claims before irreversible actions, and act on the verdic…

verify_before_acting

Pre-commitment check for one claim: picks the cheapest adequate tier (quick for stable knowledge/logic, verify for current-world claims), then applies the verdict policy — refuted = hard stop, inconcl…

How to use

Add to your Claude Desktop / Cursor / Cline MCP config:

{
  "mcpServers": {
    "second_opinion": {
      "url": "https://secondopinionx402.com/mcp",
      "transport": "streamable-http"
    }
  }
}