Loading…
Loading…
Tag
4 posts with this tag.
·9 min read
Copy-paste TypeScript patterns for AI agent logging and audit trails: structured span logging with OpenTelemetry, PII-safe trace storage, compliance-ready audit records, token usage tracking, and decision trail for human review. Working code.
·9 min read
Copy-paste TypeScript patterns for AI agent output validation: Zod schema enforcement, PII redaction, content policy filtering, JSON repair, hallucination guardrails, and cost circuit breakers. Working code.
·14 min read
Paste-ready TypeScript for tool allowlists, scoped tokens, human-in-the-loop gates, rate limits, audit logging, and RBAC. 6 working patterns for Express or Next.js, add authorization to your agent in under 20 minutes.
·14 min read
Incident response playbook for production TypeScript AI agents: 4 drop-in modules (injection detection, circuit breaker, audit logger, tool authorization) plus a decision matrix for handling an agent breach. Express and Next.js ready, full Vitest suite, working in under 30 minutes.