Context-Window Research
Empirical research on context arrangement, noise, and extraction quality in AI workflows.
Within-corpus findings, not a universal rule for context or model quality.
Research and building
A preprint, developer tools, and a safety prototype. Each establishes something different; the limitations belong alongside the artifact.
Empirical research on context arrangement, noise, and extraction quality in AI workflows.
Within-corpus findings, not a universal rule for context or model quality.
A Git hook and MCP server that checks registered review prerequisites before a mediated commit.
Registered prerequisites are not proof that a review happened or covered the current changes.
A safety research prototype that compares an agent's stated plan with its recorded actions.
Constructed examples; real-world effectiveness is not established.
A VS Code extension that brings multiple model perspectives, explicit dissent, and follow-up into one developer workflow.
Installable software; decision-quality gains and sustained external use are not established.