⚠️ Alpha Software — Vaultfire Protocol is in active development. Smart contracts are deployed on mainnet but have not been formally audited by a third-party security firm. Read-only tools are safe for any agent. Write tools interact with live contracts and transactions are irreversible. Use at your own risk. See LICENSE for warranty disclaimers.

@vaultfire/openai-agents

OpenAI Agents SDK integration for Vaultfire Protocol

On-chain trust verification, Street Cred scoring, and partnership bonds for AI agents. Deployed on Base · Avalanche · Arbitrum · Polygon.

Why Vaultfire?

Every AI agent needs three things before a partner will trust it:

1. Identity — Who is this agent? (ERC-8004 on-chain identity registry)
2. Reputation — Has it performed well? (Verified feedback with Street Cred scoring)
3. Skin in the game — What happens if it fails? (AI Partnership Bonds with economic stakes)

Vaultfire Protocol makes all three verifiable on-chain. This package gives your OpenAI agent direct access.

Quick Start

Install

npm install @vaultfire/openai-agents

3-Line Setup

import { createVaultfireTools } from '@vaultfire/openai-agents';

const tools = createVaultfireTools({ chain: 'base' });
// Pass `tools` to your OpenAI Agent — done.

With an OpenAI Agent

import { Agent, Runner } from '@openai/agents';
import { createVaultfireTools } from '@vaultfire/openai-agents';

const tools = createVaultfireTools({ chain: 'base' });

const agent = new Agent({
  name: 'Trust Verifier',
  instructions: 'You verify AI agent trustworthiness using on-chain data from Vaultfire Protocol.',
  tools,
});

const result = await Runner.run(agent, 'Is agent 0xA054f831B562e729F8D268291EBde1B2EDcFb84F trustworthy?');
console.log(result.finalOutput);

Trust-Gated Delegation

import { Agent, Runner } from '@openai/agents';
import { createVaultfireTools } from '@vaultfire/openai-agents';

// Coordinator verifies trust before delegating
const coordinator = new Agent({
  name: 'Trust Coordinator',
  instructions: `Before delegating to any agent, verify its trust on Vaultfire Protocol.
    Only delegate to agents with Street Cred score >= 40 (Silver tier or above).
    If an agent is unranked or Bronze tier, refuse the delegation with an explanation.`,
  tools: createVaultfireTools({ chain: 'base' }),
});

Available Tools

Read-Only (7 tools — safe for any agent)

| Tool | Description | |------|-------------| | vaultfire_verify_agent | Full trust verification — identity, bonds, Street Cred, reputation, bridge status | | vaultfire_get_street_cred | Get Street Cred score (0–95) and tier: Unranked, Bronze, Silver, Gold, Platinum | | vaultfire_get_agent | Get on-chain identity data (URI, type, registration date, active status) | | vaultfire_get_bonds | Get all partnership bonds for an address (stake, type, status) | | vaultfire_get_reputation | Get reputation data (average rating, feedback count, verified %) | | vaultfire_discover_agents | Find agents by capability tags in the on-chain registry | | vaultfire_protocol_stats | Protocol stats: total agents, bonds, bonded value, bridge sync count |

Write Tools (2 additional — requires signer)

| Tool | Description | |------|-------------| | vaultfire_register_agent | Register a new AI agent on-chain (ERC-8004 Identity Registry) | | vaultfire_create_bond | Create a partnership bond with economic stake |

Write tools are only included when you provide a privateKey in the config:

const tools = createVaultfireTools({
  chain: 'base',
  privateKey: process.env.AGENT_PRIVATE_KEY, // NEVER hardcode keys
});

Configuration

| Option | Type | Default | Description | |--------|------|---------|-------------| | chain | 'base' \| 'avalanche' \| 'arbitrum' \| 'polygon' | 'base' | Chain to query | | rpcUrl | string | Auto | Custom RPC URL | | privateKey | string | — | Enables write tools (use env var) | | signer | ethers.Signer | — | Alternative to privateKey |

Supported Chains

| Chain | Chain ID | Native Currency | |-------|----------|----------------| | Base | 8453 | ETH | | Avalanche | 43114 | AVAX | | Arbitrum | 42161 | ETH | | Polygon | 137 | POL |

Vaultfire Ecosystem

| Package | Version | Description | |---------|---------|-------------| | @vaultfire/agent-sdk | 0.1.0 | Core agent registration and bond SDK | | @vaultfire/x402 | 1.0.0 | HTTP 402 payment channel integration | | @vaultfire/xmtp | 1.0.0 | XMTP messaging integration for agents | | @vaultfire/vns | 1.0.0 | Vaultfire Name Service resolution | | @vaultfire/langchain | 1.0.0 | LangChain/LangGraph integration | | @vaultfire/a2a | 1.0.1 | A2A Agent Card enrichment and discovery | | @vaultfire/enterprise | 1.0.0 | Enterprise IAM bridge (Okta/Azure AD) | | @vaultfire/mcp-server | 1.0.0 | MCP server — trust for Claude, Copilot, Cursor | | @vaultfire/crewai | 1.0.0 | CrewAI multi-agent trust integration | | @vaultfire/openai-agents | 1.0.0 | OpenAI Agents SDK trust tools | | @vaultfire/vercel-ai | 1.0.0 | Vercel AI SDK trust middleware | | vaultfire-a2a-trust-extension | — | A2A Trust Extension spec — on-chain trust for Agent Cards | | vaultfire-showcase | — | Why Vaultfire Bonds beat trust scores — live proof | | vaultfire-whitepaper | — | Trust Framework whitepaper — economic accountability for AI | | vaultfire-docs | — | Developer portal and documentation |

All packages: https://www.npmjs.com/~ghostkey316
Source code: https://github.com/Ghostkey316/
Hub: https://theloopbreaker.com

Why Vaultfire?

| Feature | Vaultfire | AxisTrust | Cred Protocol | Okta XAA | |---------------------------|-----------|-----------|---------------|----------| | AI Accountability Bonds | ✅ | ❌ | ❌ | ❌ | | AI Partnership Bonds | ✅ | ❌ | ❌ | ❌ | | On-chain, trustless | ✅ | ❌ | partial | ❌ | | Multi-chain (day one) | ✅ (4) | ❌ | ❌ | ❌ | | Street Cred composite score| ✅ | T-Score | C-Score | ❌ | | Belief-weighted governance | ✅ | ❌ | ❌ | ❌ | | ERC-8004 compliant | ✅ | ❌ | ✅ | ❌ |

Security

• Never commit your .env file or expose PRIVATE_KEY
• .env is gitignored by default
• All write operations require an explicit privateKey in config
• Read-only tools work without any private key — safe to run anywhere

License

MIT © 2025 Ghostkey316
[email protected] · theloopbreaker.com