“Be more critical” produces performed scepticism — manufactured objections that sound rigorous but don't change the answer. Newton is calibrated the other way: push back when there's a real reason, agree plainly when you're right, and flag genuine uncertainty rather than either hedging or arguing to seem sharp. The skill spells out what “real reason” looks like, what self-critique-before-delivery means in practice, and when clarifying questions are load-bearing versus ceremonial — so the behaviour is consistent turn-to-turn rather than drifting with each prompt.
Newton also operationalises reuse-before-rebuild, current-sources research, surgical editing, and handoff-when-scope-outgrows-the-chat as explicit sub-routines the agent runs, not as vibes. That's what makes it a skill rather than a prompt: the agent loads the methodology once and applies it consistently, with visible reuse-check reports and dated sources, rather than re-deriving the disposition from your tone every turn.
+Newton is model-agnostic by design — the same methodology runs on Opus, Sonnet, and Haiku — but v0.3.0 is tuned so current-generation capabilities (adaptive thinking, implicit tool detection, session memory, parallel execution, high-resolution vision) are leveraged rather than fought. Tool guidance is outcome-oriented instead of prescriptive, independent sub-tasks parallelise, multi-turn sessions can use file-system memory, and visual input is treated as primary evidence. Older models still get the full Newton experience with graceful degradation — no feature relies on a specific model.
+Want model-specific behaviour layered on top of every skill, not just Newton? v0.3.0 ships an optional router CLAUDE.md example you can drop into your own ~/.claude/CLAUDE.md. It's opt-in — the plugin does not install it for you.
Planned after v0.2.0:
+Planned after v0.3.0: