Archive
Historical snapshots of the AI tool stack, rate limits, and infrastructure decisions.
Lab
Lab
Dated build logs, security investigations, and historical stack notes. Read published findings separately from experiments with results still pending.
The Lab
Where we document work in progress: current tooling, security investigations, and lessons learned the hard way.
Unlike polished guides, these are field notes—raw, evolving, and honest about what broke.
Quick Navigation
| Section | What You’ll Find | Updated |
|---|---|---|
| Now → | Current stack, active tools, what’s blocked | Continuous |
| Investigations → | Security research, experiments in progress | Sporadic |
| Archive → | Historical snapshots, how things changed | Dated entries |
Current Stack Snapshot
Read the dated stack notes in their original context. For model and tool decisions, start with the model guide and tool directory, checking their review dates.
Investigations
LLM App-Hacking Field Test: $1,500 Takeaways — A cautious read of an informal June 2026 app-security agent test.
Toybox #001: Free Zen Clock Protocol — Protocol preview—results pending; seven currently listed-Free Zen routes and 21 official attempts are locked for a later owner-run release.
Active research stays off the published site until the draft is promoted to a public lab note or a post.
Recent Archive Entries
- 11 Feb 2026 — Kimi quota resolved, Codex solid, Opus deadline tax
- 9 Feb 2026 — Rate limit reality check
- 4 Feb 2026 — OpenCode limits cut, Kiro CLI changes
Lab notes are dated and may become outdated. Check published guides for current best practices.
Investigations
Security research, experiments in progress, and deep dives.
Now
Lab setup last observed August 11, with September 27 model-selection links. Personal usage has not been revalidated.