Technical Essays & Archive
Technical write-ups on building software in public — architecture, CI, agents, and lessons from the repos I ship.
- #agent-brain
agent-brain: the architectural pivot — from passive retrieval to deterministic cognitive runtime
Part 3 of the agent-brain series: why passive context retrieval broke down in real-world autonomous coding, how we pivoted into a hard-gated execution runtime with token budgets, verify invariants, and upstream MCP routing.
- #agent-brain
agent-brain: memory scopes, temporal pruning, and cross-session routing
Part 2 of the agent-brain series: how we evolved local MCP routing with global vs project memory scopes, temporal fact decay, cross-session ingest, and circuit-broken upstream tools.
- #agentrun
AgentRun: Building Isolated, Verifiable Coding Agent Environments
The architecture and ongoing implementation of AgentRun: sandboxing coding agents with Apple containers, git worktree isolation, append-only decision logging, and bounded repair loops.
- #dustmap
Building DustMap: A Blazing-Fast, Transparent macOS Disk Analyzer in Rust and Tauri
Why I built DustMap: parallel Rayon filesystem traversal, flat memory indexes with zero pointer chasing, APFS dual-metric reconciliation, and an explainable Trash-first cleaner.
- #agent-brain
agent-brain: local MCP routing under a token budget
How I built a Rust MCP server that routes each agent turn to the right skills, rules, and memory — locally, with hook-enforced route_task and strict token limits.
Starting this dev journey blog
Why I set up a personal blog with RSS, and what I plan to write about.