Restoring the Models on a Clean Mac
2026-10-23
Part 8 of Oracles, Traces, Triage. Resetting a Mac, restoring eight local model profiles, and separating a copied directory from a verified backup.
Blog
Local Models and Tool Permissions
2026-10-22
Part 7 of Oracles, Traces, Triage. The concern that came up while choosing community models, and the tool permissions I need to settle before using them for repository investigations.
Testing Longer Contexts
2026-10-21
Part 6 of Oracles, Traces, Triage. Moving beyond an 8K default with measured inputs, explicit caps, follow-up requests, and rejection tests.
Self-Hosted Inference on Apple Silicon
2026-10-20
Part 5 of Oracles, Traces, Triage. Returning to the agent experiments with a local model server: what runs on my Macs, and what needs testing before I can use it for that work.
Agents as Fuzzers
2026-02-16
Part 4 of the Oracles, Traces, Triage series. A fuzzer is a search tool whose results must be triaged. An AI agent is a search tool whose results must be triaged. Perhaps that's not a coincidence.
Testing the Bugs Between Calls
2026-02-15
Part 3 of the Oracles, Traces, Triage series. On combining agent swarms and agent skills with stateful property-based testing to find the consensus bugs that live in sequences, not single calls.
Agent Teams and claude-swarm
2026-02-08
Part 2 of the Oracles, Traces, Triage series. On running multiple agents in parallel through git—no orchestrator, no message passing—and why the pattern matters more than the tool.