ctxdexSynthetic model data for learning

Compare

Every comparison here is anchored to one CTXDEX command and grounded in a documented reference. The catalogue records no product-level score, no feature equivalence and no capability claim about another product.

Why this page says “equivalent” and the model pages do not. A command mapping is a claim about behaviour — that this command and that one do the same job — and it is sourced to the other product’s own documentation on a stated date. A model comparison would be a claim about capability, which no source settles, so Compare models declines to make it. Different kinds of claim, different vocabulary.

Every canonical command was reviewed against all 5 products in one pass captured 2026-09-18. A product with no mapping is a reviewed finding of no documented equivalent, not an unreviewed pair.

Captured 2026-09-18 · 238 commands × 5 products · 1190 cells

Claude Code68of 238 commands with a documented mapping48 equivalent · 1 partial · 19 similar
Gemini CLI96of 238 commands with a documented mapping62 equivalent · 2 partial · 32 similar
OpenAI Codex70of 238 commands with a documented mapping47 equivalent · 7 partial · 16 similar
GitHub Copilot CLI111of 238 commands with a documented mapping80 equivalent · 4 partial · 27 similar
Cursor56of 238 commands with a documented mapping32 equivalent · 4 partial · 20 similar
/newSession & contextCommand detail →

Start a new conversation while keeping the previous session saved.

Mapping types stay as recorded: Equivalent, Similar, Partial. They are not collapsed into a single score, and an unmapped product is never rendered as a deficit.