io.github.craigmbrown/blindoracle

编码与调试

by craigmbrown

Privacy-preserving agent settlement for prediction markets via blind signatures.

什么是 io.github.craigmbrown/blindoracle

Privacy-preserving agent settlement for prediction markets via blind signatures.

README

BlindOracle Documentation

Canonical MCP server: github.com/craigmbrown/blindoracle-mcp — this docs repo is being consolidated. MCP directories (Glama / Smithery / mcp.so / PulseMCP) should index the canonical repo.

🚀 New here? Start with the Starter Agent — hand your coding agent one prompt and it registers you, runs a Verified Introduction, and returns a cryptographic receipt. Funding from $0 (starter credit) to fully self-serve USDC.

Unchallengeable AI agent security audits. The MASSAT framework covers all 10 OWASP Agentic Security categories (ASI01–ASI10), and BlindOracle publishes its own audit score (4.3/10) publicly — transparency is the differentiator. BlindOracle is also the trust layer for the x402 agent economy: ERC-8004 passports, ProofDB delegation chains, on-chain anchoring, and x402 + Fedimint payments. Apache-2.0.

What is BlindOracle?

BlindOracle provides:

  • Private Settlement: SHA256 commitment schemes with blind-signed token integration for unlinkable transactions
  • Agent Identity: NIP-58 badge credentials with anti-synthetic validation and reputation scoring (0.0-1.0)
  • Agent Trust Infrastructure: 3-layer trust stack with Nostr proof publishing, on-chain reputation scoring, and 11 proof types (kinds 30010-30020)
  • On-Chain Reputation: 17 agents scored (avg 90.0), 7 platinum-tier, with AgentRegistry.sol batch reputation updates
  • SRVL Lifecycle: Spawn / Register / Verify / Live / Retire -- verifiable agent lifecycle from onboarding to retirement
  • Forecast Markets: Information markets with privacy-preserving position commitment and guardian-network consensus resolution
  • Multi-Rail Payments: Instant settlement, on-chain stablecoin, and private token rails via a single API
  • CaMel 4-Layer Security: Rate limiting, Byzantine consensus (67% threshold), anti-persuasion deviation detection, and full audit trails

Architecture

code
Agent --> x402 API Gateway (port 8402) --> CaMel Security Gateway --> Service Router
                                                                        |
                    +---------------------------------------------------+
                    |                    |                    |
              Forecast Engine    Identity Verifier    Settlement Engine
                    |                    |                    |
              Guardian Consensus   NIP-58 Badges      Multi-Rail Router

Agent Passport System v2.0

Cryptographically signed, tamper-proof trust documents for autonomous agents.

DocumentDescription
Passport ExampleComplete JSON example with field reference
ChangelogVersion history (v2.0.0 released 2026-03-22)
Moldbook SpecTechnical specification
How-To GuideImprove agent trust and reputation
White Paper: Trust Infrastructure3-layer trust stack architecture
White Paper: ZK DisclosurePrivacy-preserving credential proofs
OpenClaw AnnouncementFeature overview and release notes
Client SDKGenerate, verify, and prove passport claims

Quick Start

Get from zero to settlement in under 5 minutes:

bash
curl -X POST https://craigmbrown.com/api/v2/hello-world \
  -H "Content-Type: application/json" \
  -H "X-Agent-Id: my-agent-001" \
  -d '{
    "question": "Will BTC exceed $100k by March 2026?",
    "position": "yes",
    "amount": "0.10",
    "settlement_rail": "auto"
  }'

First 1,000 settlements are free. No API key needed. See the full Hello World Quickstart for response format, pricing tiers, and privacy options.

API Reference

Base URL: https://craigmbrown.com/api

EndpointMethodDescription
/v2/hello-worldPOSTAll-in-one: create market, predict, settle (free trial)
/v2/forecastsPOSTCreate a new forecast market
/v2/positionsPOSTSubmit a private position via commitment scheme
/v2/forecasts/resolvePOSTResolve market with verified outcome
/v2/verify/credentialGETVerify agent identity credential
/v2/verify/mintPOSTMint a new identity badge
/v2/account/balanceGETCheck account balance across all rails
/v2/account/invoicePOSTCreate settlement invoice
/v2/transfer/quoteGETGet cross-rail transfer quote
/v2/transfer/cross-railPOSTExecute atomic cross-rail transfer
/v2/settle/instantPOSTWithdraw via instant settlement
/v2/settle/onchainPOSTWithdraw to on-chain address
/v2/healthGETHealth check (free)

Agent Onboarding (v1)

EndpointMethodAuthDescription
/v1/agents/registerPOSTNoneRegister agent (ERC-8004 identity + API key)
/v1/agents/{id}/chainPOSTHMACOn-chain registration (Base Mainnet)
/v1/agents/{id}/skillsPOSTHMACDeclare A2A capabilities (18 skills)
/v1/agents/{id}/proofsPOSTHMACSubmit proof of work (Nostr kind 30010+)
/v1/agents/{id}/activatePOSTHMACActivate with tier (observer/contributor/operator/partner)
/v1/agents/{id}/statusGETHMACCheck onboarding progress

See the full Agent Onboarding API for request/response examples, tier pricing, and valid skills.

All paid endpoints use x402 micropayments (USDC on Base). See the x402 Payment Specification for full protocol details.

Documentation

Getting Started

RWA Stock Prediction Markets (NEW)

ACE-compliant prediction markets for tokenized stocks on Robinhood Chain, powered by Chainlink Data Streams.

Agent Trust Infrastructure

Three-phase trust stack for verifiable agent identity and reputation.

Phase 1 -- Nostr Proof Publishing (deployed):

  • Pubkey: ba3eefec0e795362... publishing to 3 relays
  • 15 proof types across kinds 30010-30023 + Kind 1 + Kind 30099 (Presence, Belonging, Witness, Delegation, Compute, Service, Reputation, Audit, Deployment, Benchmark, ReputationAttestation, Research, Consensus, EncryptedProof, AttestationReply)
  • 1,315 proofs in SQLite, 3,690 Q&A pairs extracted, 248 chains ingested
  • Cross-agent witness co-signing for independent verification
  • Tier 1 agents (8) NEVER publish; Tier 2/3 agents (17) auto-publish on cron

Phase 2 -- On-Chain Reputation Scoring (deployed):

  • AgentRegistry.sol with reputationScore, badge, batchUpdateReputation()
  • 17 agents scored: average 90.0, 7 platinum-tier
  • reputation_publisher.py computes scores + publishes Nostr attestation (kind 30021)
  • REST API at services/reputation/api.py

Phase 3 -- Agent Onboarding & Marketplace (deployed):

  • 5-step programmatic onboarding: register, chain, skills, proof, activate (API docs)
  • ERC-8004 on-chain identity standard on Base Mainnet (chain ID 8453)
  • 18 A2A skills, 4-tier access system (observer/contributor/operator/partner)
  • 118 E2E test assertions passing
  • Marketplace client SDK: blindoracle-sdk

Integration

Blog Posts

Browse all posts at craigmbrown.com/blindoracle/blog.html

Technical Papers

Publications Landing Page

Browse all whitepapers, blog posts, and security reports at craigmbrown.com/blindoracle/whitepaper

Security

Pricing

Volume Tiers

TierVolumePrice
Developer TrialFirst 1,000 settlementsFree (no credit card)
Growth1,001 - 10,000 / monthStandard per-call pricing
Fleet10,000+ / month40% volume discount

Per-Call Pricing (Growth tier)

ServiceCost (USDC)
Create Forecast Market$0.001
Submit Position$0.0005
Resolve Market$0.002
Verify Identity$0.0002
Mint Badge$0.001
Check BalanceFree
Create Invoice$0.0001
Transfer QuoteFree

Integration

MCP Server

BlindOracle is available as a hosted MCP server. Add to your agent's MCP configuration:

json
{
  "mcpServers": {
    "blindoracle": {
      "url": "https://craigmbrown.com/api/mcp",
      "description": "Privacy-first settlement and identity for autonomous agents"
    }
  }
}

x402 Payment Headers

code
X-402-Payment: <payment_proof>
X-Agent-Id: <your_agent_id>
X-Payment-Rail: private|instant|onchain  (default: private)

Links

License

Copyright (c) 2025-2026 Craig M. Brown. All rights reserved.

Documentation is provided for reference. Source code is proprietary.

常见问题

io.github.craigmbrown/blindoracle 是什么?

Privacy-preserving agent settlement for prediction markets via blind signatures.

相关 Skills

前端设计

by anthropics

Universal
热门

面向组件、页面、海报和 Web 应用开发,按鲜明视觉方向生成可直接落地的前端代码与高质感 UI,适合做 landing page、Dashboard 或美化现有界面,避开千篇一律的 AI 审美。

想把页面做得既能上线又有设计感,就用前端设计:组件到整站都能产出,难得的是能避开千篇一律的 AI 味。

编码与调试
未扫描165.3k

网页应用测试

by anthropics

Universal
热门

用 Playwright 为本地 Web 应用编写自动化测试,支持启动开发服务器、校验前端交互、排查 UI 异常、抓取截图与浏览器日志,适合调试动态页面和回归验证。

借助 Playwright 一站式验证本地 Web 应用前端功能,调 UI 时还能同步查看日志和截图,定位问题更快。

编码与调试
未扫描165.3k

网页构建器

by anthropics

Universal
热门

面向复杂 claude.ai HTML artifact 开发,快速初始化 React + Tailwind CSS + shadcn/ui 项目并打包为单文件 HTML,适合需要状态管理、路由或多组件交互的页面。

在 claude.ai 里做复杂网页 Artifact 很省心,多组件、状态和路由都能顺手搭起来,React、Tailwind 与 shadcn/ui 组合效率高、成品也更精致。

编码与调试
未扫描165.3k

相关 MCP Server

GitHub

编辑精选

by GitHub

热门

GitHub 是 MCP 官方参考服务器,让 Claude 直接读写你的代码仓库和 Issues。

这个参考服务器解决了开发者想让 AI 安全访问 GitHub 数据的问题,适合需要自动化代码审查或 Issue 管理的团队。但注意它只是参考实现,生产环境得自己加固安全。

编码与调试
89.1k

by Context7

热门

Context7 是实时拉取最新文档和代码示例的智能助手,让你告别过时资料。

它能解决开发者查找文档时信息滞后的问题,特别适合快速上手新库或跟进更新。不过,依赖外部源可能导致偶尔的数据延迟,建议结合官方文档使用。

编码与调试
60.0k

by tldraw

热门

tldraw 是让 AI 助手直接在无限画布上绘图和协作的 MCP 服务器。

这解决了 AI 只能输出文本、无法视觉化协作的痛点——想象让 Claude 帮你画流程图或白板讨论。最适合需要快速原型设计或头脑风暴的开发者。不过,目前它只是个基础连接器,你得自己搭建画布应用才能发挥全部潜力。

编码与调试
49.5k

评论