Earned Autonomy
Thursday, 9am At nine on Thursday morning a container started in Azure, pulled the week's sources, clustered the same story where several outlets had covered...
Measurement tools need evals too
A tool that refuses to answer Here is one of my tools declining to do the thing it exists to do. $ git-habits scan --from-export thin.tsv selection window...
One LLM judge is an opinion. Two model families agreeing is evidence.
The problem with one judge I migrated my news pipeline's article rewriting from a local Qwen model to hosted Claude (previous piece), and before cutting over...
The migration the data ordered: moving a production LLM pipeline off the Mac Mini under my desk
The bill, up front Last week my news pipeline rewrote a full weekly batch of 85 articles in the cloud, across all four of its writer personas. The rewriting...
An MCP server that has to prove itself: citations, abstention and entitlements as CI gates
The demo, in one screen Here's the eval scorecard for a small MCP server on a green build: retrieval (staff, n=22): hit@1 1.00 recall@5 1.00 MRR 1.00...