Smeldr separates AI agent proposals from human-ratified decisions
The author argues that AI coding agents struggle with unclear decisions, since notes show what was written but not which decision is in force or who approved it. Smeldr addresses this by keeping context and references separate from a governed set of currently binding decisions, each with its lineage and traceability. Agents can propose decisions, but only a person with ratification rights can adopt them, after which they become immutable. Changes are made by superseding an old decision with a new one, which keeps the original on record. In a demonstration, an agent's attempt to ratify was refused by the server, and when a dependent decision rested on a superseded one, the system flagged it for review.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.
Discussion (0)
Log in to join the discussion and vote.
Log in