_ __ __ _____
_ __(_)___ / /____ __________/ /_____ /__ /__
| | /| / / / __ \/ __/ _ \/ ___/ ___/ __/ __ `/ / / _ \
| |/ |/ / / / / / /_/ __/ / (__ ) /_/ /_/ / / / __/
|__/|__/_/_/ /_/\__/\___/_/ /____/\__/\__,_/ /_/\___/
Building tools for AI coding agents, and using every new framework I can get my hands on to do it.
ββ ~ ββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β $ whoami β
β agentic coding Β· AI tooling Β· one new stack per project β
β β
β $ uname -a β
β WSL2 Β· Docker Β· a shell with an agent running in it β
β β
β $ ls ~/projects β
β agentdeck/ augur/ nexum/ Doc-Summarizer/ β
β FlowForge/ AiGameCompanion/ β
β β
β $ alias ship β
β ship='agent | lint | typecheck | test && git push' β
β # a failed stage goes back to the agent, never to main β
β β
β $ cat .rules β
β local-first Β· no cloud Β· no telemetry β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
| Project | What | Stack |
|---|---|---|
| agentdeck | Desktop deck for WSL coding agents β 7 agents + custom CLIs, split terminals, visual workflows, worktree isolation | Electron Β· React Β· TS |
| augur | Local-first neurosymbolic agent β blackboard over Redis + NATS, gated local-LLM advice, self-tuning | Python Β· Ollama |
| nexum | Cross-agent memory for Claude Code + Codex CLI β signed, searchable, served over MCP | Rust Β· SQLite |
| Doc-Summarizer | Fully offline summarizer β map-reduce over long docs, every point traced to its source sentence | Python Β· llama.cpp |
| FlowForge | Node-based file pipelines β rename, resize, convert, filter β GUI and CLI over one engine | C# Β· .NET 10 Β· Avalonia |
| AiGameCompanion | In-game AI overlay drawn straight into DX11/DX12 titles | Rust Β· hudhook |
Personal tools, not products. Open source, no support guarantees β issues and PRs are read.

