What problem does it solve? Voice speaking practice done in the Claude app's Fluent Speaking project produces results outside the Fluent tracking system, so scores, mistakes, vocabulary, and streaks would be lost unless manually re-entered. This Skill closes that loop by fetching, validating, and recording the session exactly like a native speaking session. ## Core Features & Use Cases - Multi-source result retrieval: Pulls the results JSON from pasted text, a Slack channel, Google Drive, or reconstructs scores from a raw chat transcript. - Authoritative re-scoring: Validates every field, clamps score ranges, recomputes totals, relabels error categories, and blocks duplicate or cross-language imports. - Full database integration: Converts results into a single update-db.py payload that writes the session markdown, logs mistakes, schedules SM-2 reviews, and updates the streak. - Use Case: After finishing a voice practice session in the Claude app, say "import my speaking results" and the Skill fetches the JSON from Slack, re-scores it, and updates all six Fluent databases. ## Quick Start Tell the assistant "I finished my voice practice, pull in my speaking results" after completing a voice session created by /fluent-speaking-export.
