fluent-session-analyzerlisted

# Session Analyzer ## Overview Every practice session writes a markdown report to `/results/fluent-{skill}-session-{NNN}.md` (e.g. `fluent-writing-session-012.md`). This skill describes how to read those files to plan adaptive follow-up practice. Use it when the tutor needs narrative context the JSON databases don't capture — the exact sentence the learner wrote, the scenario, the feedback they received. ## When to Use Load this skill whenever the tutor: - Plans today's focus before `/fluent-learn`, `/fluent-writing`, etc. - Answers the learner's question "what's my weakest area" or "what should I work on". - Generates the next session plan. Skip this skill when aggregated JSON numbers are enough — prefer `read-db.py` for counts, trends, and mastery levels. Use this skill only when the textual context matters. ## Instructions ### 1. Find recent session files ``` /results/fluent-{skill}-session-{NNN}.md ``` `NNN` is the global session counter, so files group by skill and sort chronologically. Files written before v0.2.0 lack the `fluent-` prefix (`{skill}-session-{NNN}.md`) — glob for both, and don't rename the old ones. Read the most recent 3-5 files of the relevant skill; don't re-read the entire history. ### 2. Extract error patterns Scan for `❌` markers. Each correction has: - The wrong form ("Your answer") - The correct form - A category (grammar, formal_informal, vocabulary, prepositions, articles, spelling, missing) - A severity (🔴 critical, 🟡 moderate, ���