Blog / Long form / 08
Architecture write-ups, release notes and the reasoning behind the decisions.
Short one-idea observations live in Field Notes →
Eleven teams, one benchmark, the same model families. The best and worst systems running identical models finished 30.4 points apart, and everything that opened that gap sat outside the weights.
Aug 18, 2026 · 12 min read
Read article →A model launch is usually three bars on a chart. This one is more interesting for where the improvement comes from — and for what...
Aug 13, 2026 · 12 min read
When a coding agent fails we rarely change the model. We change the harness around it — and the agent works better next time. So w...
Aug 11, 2026 · 10 min read
| No. | Article | Type | Topic | Date | Read |
|---|---|---|---|---|---|
| 04 | The wrong OpenCode session | Changelog | context-bridge | Aug 6, 2026 | 2 min |
| 05 | The first switch into OpenCode | Changelog | context-bridge | Aug 6, 2026 | 3 min |
| 06 | The day an agent ran clean in the wrong directory | Post | context-bridge | Aug 5, 2026 | 3 min |
| 07 | context-bridge 0.11 and 0.12: OpenCode, safer handoffs, and lanes | Changelog | context-bridge | Aug 5, 2026 | 7 min |
| 08 | Building context-bridge with context-bridge | Post | context-bridge | Aug 5, 2026 | 4 min |