Daniel Tse
Daniel Tse is an individual principal engineer and solo open-source developer who builds AI tools, backend systems, JVM benchmarks, and cloud-delivery systems, and publishes the code, benchmarks, and trade-offs behind them. His stated focus is making architecture claims survive implementation detail, benchmark constraints, operational guardrails, and reuse by real teams.
At a Glance
- AI tooling and agent-platform teams
- Application-security, DevSecOps, and security-automation teams
- Java/Spring backend and platform teams
- AdTech and high-throughput backend engineering teams
- +1 more
AI Tools by Daniel Tse
(1)
MCP ZAP Server
MCP Server for ZAP Security Scans
Discussions
No discussions yet
Be the first to start a discussion about Daniel Tse
Latest News
MCP ZAP Server v0.16.0 released with scanner-aware findings and CI baselines, whole-plan Automation Framework validation, safer plan/report paths, and preserved root-alert evidence in scoped reports.
MCP ZAP Server v0.15.0 released with an EKS infrastructure starter, opt-in OpenAPI content import, paged report retrieval, and stronger report/scan controls.
Published 'Loop Engineering Needs a Flight Recorder,' arguing production agent loops need replayable evidence rather than only better automation.
Published 'Demos Are Cheap. Production MCP Is Expensive,' using MCP ZAP Server to discuss ownership, authorization, deployment, observability, and live-remediation requirements.
Products & Services
Apache-2.0-licensed, self-hosted MCP server that lets AI agents and operators import API schemas, run OWASP ZAP workflows, and retrieve structured findings. It provides API-key/JWT authentication, tool scopes, policy, URL validation, quotas, limits, audit events, Docker Compose, Helm, CI security-gate examples, authenticated crawling, Client Spider, and governed report access.
Embeddable Java governance contracts for MCP tool runtimes, covering tool identity, authorization/access rules, policy decisions, audit events, abuse protection, URL scoping, correlation IDs, quotas, and rate limits, plus an optional Spring WebFlux adapter published to Maven Central.
AdTech real-time-bidding ingress benchmark comparing Quarkus JVM/native, Go, Rust, Python, Spring, and Node across HTTP-only and Kafka-backed workloads, with documented workload and fairness controls.
Python/FastAPI flight recorder for AI answer loops, with evidence routing, verification, retries, refusals, durable local run reports, JSONL inspection, evals, trace exports, local threads, indexed files, saved loop recipes, and a loop timeline.
Market Position
Daniel Tse positions the work as inspectable, production-oriented open-source infrastructure rather than toy AI integrations: MCP ZAP Server combines MCP agent access with OWASP ZAP scanning, authentication, policy, URL validation, limits, auditability, deployment artifacts, and CI evidence. MCP Gateway Core is deliberately a small embeddable governance-contract layer rather than a gateway runtime, scanner, proxy, or data plane. This differentiates the portfolio from lightweight MCP wrappers and from full commercial DAST/gateway platforms by emphasizing self-hosting, composable Java contracts, and evidence-backed operational guardrails.
Leadership
Executive Team
Daniel Tse
Principal Engineer; independent developer/creator
Builds AI-facing tools, backend platforms, JVM benchmarks, cloud systems, and technical writing. Public search results describe him as a lead/senior software engineer, ex-Sun, with more than 15 years of software-engineering experience.
Founding Story
The personal site presents Daniel Tse's work as a site-first engineering practice built around the idea that architecture needs receipts: working systems, source code, benchmarks, security boundaries, deployment paths, and operational evidence. The initial/public vision is to build AI-facing tools and backend defaults that are inspectable and useful in production rather than toy integrations.
Business Model
Revenue Model
The core MCP ZAP Server is open source under Apache-2.0. The project documents optional paid support through Agentic Lab for production adoption, SSO/OAuth2 setup, CI/CD quality gates, and compliance reporting; it also accepts maintainer sponsorship through GitHub Sponsors.
Target Markets
- AI tooling and agent-platform teams
- Application-security, DevSecOps, and security-automation teams
- Java/Spring backend and platform teams
- AdTech and high-throughput backend engineering teams
- Teams evaluating customer-facing AI systems and agent outcomes
- Security teams and DevSecOps teams running self-hosted, operator-controlled web/API security scans
- AI-agent developers needing governed MCP tool access and production security boundaries
- Teams deploying the open-source MCP ZAP Server in Docker, Kubernetes, CI/CD, or cloud environments
- Java and Spring teams embedding MCP authorization, policy, audit, abuse-protection, URL-scope, and rate-limit contracts
- Backend and AdTech engineers comparing JVM, native, Go, Rust, Python, Spring, and Node ingress behavior
- Product and AI teams simulating customer outcomes before shipping customer-facing agents