I’m an AI Engineer focused on agent operations, knowledge operations, and automation reliability.

I build automation systems and LLM/agent workflows that are not only functional, but also operable, monitorable, and maintainable in real-world environments. My specialty is tying tools, workflows, and repositories together into automation that doesn’t break.

Here are some public repos I’d recommend as entry points:

• Review / fix workflows
  • review-fix-pipeline — intent-first review and automated fix loops with independent reviewer contexts
  • claude-review-pdca — closed-loop reuse of past review findings during implementation
  • claude-code-hooks — guardrails that block user-delegation and premature completion
• Knowledge / documentation operations
  • x-bookmark-knowledge-pack — local-first X bookmark pack generator for humans and AI agents
  • doc-freshness-analyzer — verify README/docs claims against actual code reality
• Data pipeline / dashboard artifact
  • huggingface-daily-insights-api — daily snapshots, dashboard, CSV releases, and API for open AI ecosystem tracking

If you want a place to start, I’d especially recommend review-fix-pipeline for agent workflow design or x-bookmark-knowledge-pack for local-first knowledge-pack style tooling.

End-to-end view of the operations pipeline for a personal LLM Wiki. Visualizes the data flow from a Task Scheduler-triggered orchestrator, through the intake / governance / graphify flows, down to the operations cockpit.

Live demo: https://curiosity-wiki-ops.vercel.app

• 8-agent sensitivity-optimized pipeline (Phase 5)
• Concept normalization + alias resolution via a canonical layer (Phase 6)
• Automated intake 4x/day + weekly governance + daily meta supervision
• Multi-source search chain across Grok 4 / Perplexity / xAI API
• Multi-model orchestration with GLM-5 × Claude Code × Codex

The main repository is private. Only the visualization HTML is carved out into a standalone deploy repository and published via Vercel.

• curiosity-wiki — track-aware research and knowledge operations system for AI-agent-friendly workflows
  Demo: https://curiosity-wiki-ops.vercel.app
• openclaw-claude-bridge — bridge-oriented orchestration for multi-agent execution and operational routing
  Repo: https://github.com/Tenormusica2024/openclaw-claude-bridge
• vault-d — metadata normalization and knowledge structure design for durable AI workflows
  Repo: https://github.com/Tenormusica2024/vault-d

• curiosity-wiki Operations Flow (live demo)
  https://curiosity-wiki-ops.vercel.app
• openclaw-claude-bridge
  https://github.com/Tenormusica2024/openclaw-claude-bridge
• vault-d
  https://github.com/Tenormusica2024/vault-d
• Portfolio — personal portfolio site showcasing selected projects, AI workflow work, and implementation context
  https://github.com/Tenormusica2024/portfolio

• Languages: Python, TypeScript, PowerShell, Bash
• AI / Agents: Claude Code (Opus / Sonnet), Codex (GPT-5.4), GLM-5, Grok 4
• Automation: Task Scheduler, GitHub Actions, Vercel, Cloud Run
• Knowledge Ops: Obsidian Vault, frontmatter metadata, semantic graph (graphify)
• Orchestration: Claude in Chrome (CiC), Playwright CLI, dev-browser, MCP servers

• Agent operations and AI workflow reliability
• Knowledge operations and agent-friendly structure design
• Cost-aware model routing and role separation
• Safe automation that reduces manual routing and repetitive judgment
• Guardrails, monitoring, and handover for real-world operation