What problem does it solve? Developers lose track of what they accomplished across many Copilot chat sessions, making it hard to write standups, find past work, or understand how they use AI assistance. This Skill queries the Copilot session store to reconstruct activity, surface patterns, and answer questions about past sessions. ## Core Features & Use Cases - Standup Reports: Generates a formatted daily standup from the last 24 hours of sessions, grouped by branch and feature, with PR status checked via GitHub CLI or MCP tools. - Usage and Cost Tips: Analyzes prompting patterns, tool usage, token consumption (cloud DuckDB) or proxy signals like missing compactions (local SQLite) to produce personalized, data-grounded recommendations. - Session Search and Maintenance: Searches past sessions by keyword, file path, or PR/issue reference using FTS5 locally or ILIKE on cloud, plus reindexing and guidance for deleting session data. - Use Case: Before a morning standup, ask for a summary of yesterday's work and receive a grouped report of completed and in-progress features with linked PRs and session IDs. ## Quick Start Ask the assistant to generate a standup report from your Copilot session history for the last 24 hours.