feat(examples): add agent trust check via TWZRD Agent Intel MCP server

[twzrd-sol]

What

Adds examples/agent-trust-check.ts — a working example that shows how to verify an AI agent's wallet trust score (via TWZRD Agent Intel, an MCP server) before passing the result to OpenAI for a pass/fail decision.

Pattern

MCP tool call (score_agent / preflight_check) → OpenAI chat completion → ALLOW / BLOCK

This is a common multi-agent pattern: external tool → LLM decision. The TWZRD server is free for score_agent and preflight_check — no API key needed.

Usage

npm install @modelcontextprotocol/sdk
OPENAI_API_KEY=sk-... yarn tsn -T examples/agent-trust-check.ts

MCP server details

• Endpoint: https://intel.twzrd.xyz/mcp (streamable-http)
• Tools: score_agent(wallet), preflight_check(wallet) (free), get_trust_receipt(wallet) (paid x402)
• npm: twzrd-agent-intel

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 7b588eec9f

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Declare the MCP SDK dependency

This new example imports @modelcontextprotocol/sdk, but neither the root package.json nor examples/package.json/lockfile declares that package. The repo's lint path runs ./node_modules/typescript/bin/tsc with tsconfig.json including examples, so a normal checkout after yarn install/scripts/bootstrap will fail type-checking with an unresolved module before anyone has manually run the install command from this file's comment. Add the dependency to the checked-in examples package metadata so CI and consumers can resolve the import.

Useful? React with 👍 / 👎.