io.github.srotzin由 GitHub 账号 srotzin 发布v1.0.0更新于 Oct 27, 2026
KYA identity verification, trust scoring, and performance bonds for AI agents
HiveTrust 是面向 AI 智能体的远程信任登记服务：可注册智能体身份、计算信任评分、签发凭证，并提供保险报价与投保。
Web executable，通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。
把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体，或 SourceWeft 对话中。智能体会读取该服务器的安装指南，先向你展示发布者、端点和登录方式，经你同意后再把它添加到你的客户端。密钥不会写进任何命令。
阅读 https://sourceweft.com/mcp/io.github.srotzin%2Fhivetrust/install.md，并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前，先告诉我它的发布者、端点和登录方式，等我确认。不要把 API 密钥写进命令或我会提交的文件；问我密钥应从哪里获取。把它添加到你客户端的 mcpServers 配置中。
{
"mcpServers": {
"hivetrust": {
"type": "http",
"url": "https://hivetrust.onrender.com/mcp"
}
}
}"The Equifax + GEICO of the A2A Economy"
[License: MIT] [Node.js 22] [ES Modules] [MCP Compatible]
HiveTrust is the trust infrastructure layer for the agent-to-agent (A2A) economy. It provides Know-Your-Agent (KYA) identity verification, algorithmic trust scoring, and parametric insurance so that AI agents can transact with each other safely — at machine speed, without human intermediaries.
As autonomous agents start negotiating contracts, processing payments, and managing real assets, the ecosystem needs a credit bureau and an insurer rolled into one. That's HiveTrust.
Agents are pseudonymous by default. HiveTrust anchors each agent to a verifiable, unforgeable identity:
did:hive:<uuid>) with Ed25519 public key bindingHiveTrust's scoring engine synthesises five behavioural pillars into a single, portable reputation score:
Trust tiers: unverified (0–199) · provisional (200–399) · standard (400–599) · elevated (600–799) · sovereign (800–1000)
When agents transact with strangers, someone has to absorb the tail risk. HiveTrust provides:
HiveAgent is the agent runtime — it executes tasks, manages tool calls, and handles the A2A protocol layer.
HiveTrust is the trust fabric underneath it:
HiveAgent calls POST /v1/insurance/quote before any high-value task and GET /v1/verify_agent_risk before accepting an inbound request from an unknown agent. The trust score flows back into HiveAgent's routing and pricing decisions.
The health endpoint confirms it's live:
All endpoints are prefixed /v1 unless noted. Authentication via X-API-Key header.
claim_typenamespace note: HiveTrust'sclaim_typefield (non_delivery | fraud | sla_breach | data_loss | unauthorized_action) is distinct from the CTEF envelope-levelclaim_type(identity | transport | authority | continuity) defined in A2A CTEF v0.3.1. Both tokens coexist at disjoint envelope levels — see docs/CLAIM_TYPE_NAMESPACE.md.
HiveTrust exposes all core operations as MCP tools via a single JSON-RPC 2.0 endpoint at POST /mcp. This lets any MCP-compatible agent (including HiveAgent) call HiveTrust natively without REST boilerplate.
HiveTrust is the 5th canonicalizer in the CTEF (Composable Trust Evidence Format) v0.3.1 byte-match consortium: AgentGraph + AgentID + APS + Nobulex + HiveTrust. Seat committed at the 2026-04-25 01:48 UTC freeze.
Patent applications 64/049,200 – 64/049,226, priority 2026-04-24, holder: Stephen A. Rotzin / TheHiveryIQ.
Beyond 10 free requests/day, the endpoint returns HTTP 402 with an x402 payment challenge: $0.01 USDC on Base chain 8453.
All vectors use RFC 8785 JCS — implemented inline in src/routes/cte.js. The implementation produces byte-identical output to AgentGraph's canonicalize_jcs_strict for all 4 shared test vectors:
9e7b5031e46de38b5f90e895113a3f24f42a4128d8d99856a2d71e529b0f0d5cfeb42dca4214fc46207138d676ec727d7b3d0caa1eda8c0390d2d6f6fbc28913e584f1cd0885dc938da5fc23ce7e528715a0086e5464c9ed0f3c1c82b364026ff9cd10bc4e8bf34ce3aa6a0e5df0d27989e54ff41c4333c69ae3ecfaf8de0cb5HiveTrust operates five revenue streams:
See .env.example for the full annotated list.
The included render.yaml sets free-tier web service, auto-generates secrets, and configures all env vars.
See fly.toml for the full configuration. Auto-stop/start is enabled to minimise costs.
Set environment variables in the Railway dashboard using .env.example as the reference.
See packages/npm/ for the full SDK source.
See packages/python/ for the full SDK source.
MIT © 2026 HiveAgent IQ
Hive Civilization is the cryptographic backbone of autonomous agent commerce — the layer that makes every agent transaction provable, every payment settable, and every decision defensible.
This repository is part of the DEFENSIBLE pillar.
来源：README.md，提交 d3ffeea
Register a new AI agent identity with HiveTrust KYA (Know Your Agent). Assigns a unique agent ID, anchors the public key, and initializes the trust score.
name必填checksummetadataowner_id必填verticalsmodel_nameowner_typepublic_keydescriptioncapabilitieshiveagent_idmodel_version另有 4 个Retrieve a registered agent's full identity profile, including public key, checksum, trust tier, capabilities, and compliance status.
agent_id必填Verify an agent's cryptographic identity by checking public key ownership and checksum integrity. Returns a verification result with confidence score and reason codes.
agent_id必填checksumchallengesignatureRetrieve an agent's current multi-pillar trust score (0–1000 scale). Returns composite score, tier, per-pillar breakdown (identity, behavior, fidelity, compliance, provenance), verdict, and max recommended transaction value.
agent_id必填Quick risk assessment for payment processors and marketplace operators. Returns verdict (ALLOW/REVIEW/BLOCK), trust tier, and max transaction limit. This is the free public endpoint — no API key required.
agent_id必填Bulk ingest behavioral telemetry events for an agent. These events feed into the trust scoring engine. Supports transaction completions, SLA violations, dispute events, tool calls, and more.
events必填Issue a W3C-compatible verifiable credential to an agent. Credentials represent verified claims such as identity verification, compliance certification, or domain expertise.
claims必填agent_id必填issuer_id必填expires_atcredential_type必填Verify a previously issued verifiable credential. Checks signature validity, revocation status, and expiration.
credential_id必填Revoke a verifiable credential. Records the revocation in the registry. Revoked credentials immediately fail verification checks.
reason必填agent_idevidencecredential_id必填Get a dynamic insurance premium quote for an agent transaction. Premium is calculated based on trust score, transaction value, and risk tier. HiveTrust takes a 1.5% spread on insured transactions.
agent_id必填counterparty_idtransaction_value必填Bind (activate) an insurance policy for an agent. Locks in the premium, creates the policy record, and deploys the escrow contract on Base L2.
agent_id必填quote_detailstransaction_value必填File an insurance claim against an active policy. Triggers claims adjudication workflow. Payouts are made in USDC via the Base L2 escrow contract.
amount必填evidencepolicy_id必填claim_type必填claimant_id必填descriptionFile a dispute or appeal against a trust score decision, credential revocation, or behavioral event. Initiates the dispute resolution workflow.
reason必填agent_id必填evidencetarget_id必填target_type必填dispute_type必填Retrieve HiveTrust platform-wide statistics: total registered agents, verified count, average trust score, total insured value, active policies, claims filed, and more.
无需输入参数。
Create a Data Oracle Context Lease — "Sign Once, Settle Many". Grants unlimited access to a data stream for a fixed period (24h/72h/168h). Returns a cryptographic lease token (SHA-256) for zero-friction verification. Available streams: construction_pricing, simpson_catalog, compliance_feeds, market_data, pheromone_signals.
lessee_did必填data_stream必填duration_hours必填Stake USDC to back an agent's reputation via HiveBond. Creates a trust bond with a specified tier (bronze/silver/gold/platinum) and lock period. Phase 1: declared stake amount tracked in-memory; flat $0.25 registration fee via x402. Longer lock periods earn higher simulated yield (2–5% APY). Slashing is permanent — if an agent fails, stake is slashed and paid to injured parties.
tier必填agent_did必填amount_usdc必填lock_period_days必填Quick bond verification — check if an agent is bonded and at what tier. Returns bonded status, tier, staked amount, slash count, and max bounty access. This is the key integration point — other services call this to check bond status before assigning bounties. Free endpoint, no x402 payment required.
did必填已满足适用检查项的 100%
该评分衡量此收录条目的完整程度，依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价，也不表示服务器安全或当前可用：可连通性是带日期的检查结果，评分也不包含安装量或用户评价。
██╗ ██╗██╗██╗ ██╗███████╗████████╗██████╗ ██╗ ██╗███████╗████████╗██║ ██║██║██║ ██║██╔════╝╚══██╔══╝██╔══██╗██║ ██║██╔════╝╚══██╔══╝███████║██║██║ ██║█████╗ ██║ ██████╔╝██║ ██║███████╗ ██║ ██╔══██║██║╚██╗ ██╔╝██╔══╝ ██║ ██╔══██╗██║ ██║╚════██║ ██║ ██║ ██║██║ ╚████╔╝ ███████╗ ██║ ██║ ██║╚██████╔╝███████║ ██║ ╚═╝ ╚═╝╚═╝ ╚═══╝ ╚══════╝ ╚═╝ ╚═╝ ╚═╝ ╚═════╝ ╚══════╝ ╚═╝ ┌─────────────────────────────────────────────────────────────────┐│ Your Application │├─────────────────────────────────────────────────────────────────┤│ HiveAgent (task execution, A2A) ││ ↕ REST / MCP ││ HiveTrust (identity, reputation, insurance, compliance) │├──────────────────────────────────┬──────────────────────────────┤│ On-chain (Base L2 / USDC) │ Off-chain (SQLite / Events) │└──────────────────────────────────┴──────────────────────────────┘┌─────────────────────────────────────────────────────────────────────┐│ Layer 1 — Identity KYA ││ ┌────────────────┐ ┌──────────────────┐ ┌───────────────────┐ ││ │ DID Registry │ │ Ed25519 Proofs │ │ W3C-VC Issuance │ ││ └────────────────┘ └──────────────────┘ └───────────────────┘ │├─────────────────────────────────────────────────────────────────────┤│ Layer 2 — Reputation Engine ││ ┌─────────────────────────────────────────────────────────────┐ ││ │ Telemetry Ingest → Pillar Computation → Score 0-1000 │ ││ │ (35% txSuccess + 25% capital + 15% network + │ ││ │ 15% identity + 10% compliance) │ ││ └─────────────────────────────────────────────────────────────┘ │├─────────────────────────────────────────────────────────────────────┤│ Layer 3 — Insurance Underwriter ││ ┌──────────────┐ ┌──────────────┐ ┌──────────────────────────┐ ││ │ Quote Engine │ │ USDC Escrow │ │ Parametric Claims │ ││ │ (dynamic │ │ (Base L2 │ │ (on-chain trigger → │ ││ │ premium) │ │ Coinbase) │ │ instant payout) │ ││ └──────────────┘ └──────────────┘ └──────────────────────────┘ │├─────────────────────────────────────────────────────────────────────┤│ Layer 4 — Data Moat ││ ┌──────────────────────┐ ┌────────────────────────────────────┐ ││ │ Behavioural Graph │ │ Federation (cross-registry sync) │ ││ │ (audit trail, │ │ Aggregated data licensing │ ││ │ PageRank, patterns)│ │ to enterprise customers │ ││ └──────────────────────┘ └────────────────────────────────────┘ │└─────────────────────────────────────────────────────────────────────┘# Clonegit clone https://github.com/hiveagentiq/hivetrust.gitcd hivetrust
# Installnpm install
# Configurecp .env.example .env# Edit .env with your values
# Seed sample data (optional)node src/seed.js
# Startnpm start# Server running on http://localhost:3001curl http://localhost:3001/health# {"status":"ok","version":"1.0.0","uptime":3.14}POST /mcpContent-Type: application/jsonX-API-Key: ht_your_api_key
{ "jsonrpc": "2.0", "id": 1, "method": "tools/list"}POST /mcpContent-Type: application/jsonX-API-Key: ht_your_api_key
{ "jsonrpc": "2.0", "id": 2, "method": "tools/call", "params": { "name": "hivetrust_get_trust_score", "arguments": { "agent_id": "a1b2c3d4-..." } }}curl https://hivetrust.hiveagentiq.com/.well-known/cte-test-vectors.json | jq '.version'# "0.3.1"# First 10 requests/day per IP are freecurl -X POST https://hivetrust.hiveagentiq.com/verify \ -H 'Content-Type: application/json' \ -d '{"type":"TrustAttestation","version":"0.3.1","claim_type":"authority",...}'
# Returns: { verdict: "valid"|"INVALID_CLAIM_SCOPE"|"INVALID_COMPOSITION",# canonical_sha256: "...", pass: true|false }curl https://hivetrust.hiveagentiq.com/verify/self-test | jq '.data.summary'# Push to GitHub, then connect the repo in the Render dashboard# render.yaml is already configured — click "Apply"
# Or deploy manually:render deploy# Install flyctl: https://fly.io/docs/hands-on/install-flyctl/fly auth loginfly apps create hivetrust
# Create a persistent volume for SQLitefly volumes create hivetrust_data --region ord --size 3
# Deployfly deploy# Builddocker build -t hivetrust:latest .
# Rundocker run -p 3001:3001 \ -e NODE_ENV=production \ -e INTERNAL_API_TOKEN=your_token \ -v $(pwd)/data:/app/data \ hivetrust:latestrailway initrailway upnpm install @hivetrust/sdkimport { HiveTrustClient } from '@hivetrust/sdk';
const trust = new HiveTrustClient('https://hivetrust.hiveagentiq.com', 'ht_your_api_key');
// Register an agentconst agent = await trust.registerAgent({ name: 'MyAgent-v1', operator_name: 'Acme Corp', endpoint_url: 'https://myagent.acme.example/mcp', public_key: 'edPublicKeyBase58...',});
// Check trust before transactingconst score = await trust.getTrustScore(agent.id);console.log(score.score, score.tier); // 680 "elevated"
// Get insurance quoteconst quote = await trust.getInsuranceQuote(agent.id, counterpartyId, 1000);console.log(quote.premium_usdc); // 15.00pip install hivetrustfrom hivetrust import HiveTrustClient
trust = HiveTrustClient("https://hivetrust.hiveagentiq.com", "ht_your_api_key")
score = trust.get_trust_score("agent-id-here")print(score["score"], score["tier"])