learning-statslisted

# /learning-stats Read-only. Makes no changes to `state.json`. May write a chart image if requested (see step 5) — that's the only file it touches. ## Steps 1. If `state.json` doesn't exist, say so and suggest running `/month`. Stop. 2. Scan **every** month in `state.json` and compute: - Total days done, total rest days, total days remaining (pending + current) across the whole project. - Total time studied: sum of `time_spent_minutes` across all done days that have it set — report it, but note plainly if some done days have `null` (self-report was skipped) so the total is visibly a lower bound, not a hidden guess. - Confidence distribution (`strong`/`shaky`/`struggled` counts) and difficulty distribution (`easy`/`medium`/`hard` counts) across all done days. - Most common `tags` across all days, with counts. - Micro-projects: total attempted vs. completed. - All `badges_unlocked`, in the order earned. - The cross-month study streak (same computation as `/progress`). 3. Present this as a compact, readable summary — not a wall of raw numbers. Group related figures together (e.g. time and pace together, confidence and difficulty together). 4. If there's a meaningful trend worth naming (e.g. confidence improving over the last several days, or time-per-day trending down as the user gets faster), say so in plain language — but only if the data actually supports it; don't manufacture a trend from a handful of points. 5. **Optional chart**: if the user ask