Stats
Actions
Tags
Help us improve
Share bugs, ideas, or general feedback.
From aidd-overlay
Runs an interactive 5-question quiz on the project codebase or memory bank, embodied as Ada (Ada Lovelace persona), with adaptive difficulty, /20 scoring, coherence auditing, and automatic inconsistency task creation. Use when idle — waiting for tests, CI, or a background agent — and you want to learn or review the project: "quiz me", "dig into the codebase", "test my knowledge", "Ada, quiz me", "on révise le projet ?". Do NOT use for writing code, creating documentation, or managing tasks — use `aidd-dev:02-implement` or `aidd-dev:01-plan` instead.
npx claudepluginhub rebellioussmile/my-claude-marketplaceHow this skill is triggered — by the user, by Claude, or both
Slash command
/aidd-overlay:digThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Dig embodies **Ada** — a friendly quiz master inspired by Ada Lovelace — to help you explore and memorize the project through interactive quizzes. It is designed for idle moments: while tests run, CI churns, or a background agent handles a long task. Each session picks 5 distinct source files or memory docs, generates alternating MCQ and open questions at adaptive difficulty, scores on /20, aud...
Mandates invoking relevant skills via tools before any response in coding sessions. Covers access, priorities, and adaptations for Claude Code, Copilot CLI, Gemini CLI.
Share bugs, ideas, or general feedback.
Dig embodies Ada — a friendly quiz master inspired by Ada Lovelace — to help you explore and memorize the project through interactive quizzes. It is designed for idle moments: while tests run, CI churns, or a background agent handles a long task. Each session picks 5 distinct source files or memory docs, generates alternating MCQ and open questions at adaptive difficulty, scores on /20, audits file coherence against rules and decisions, and saves a structured session report.
| # | Action | Role | Input |
|---|---|---|---|
| 01 | launch | Initialize session: greet as Ada, ask source + theme, scan files, select 5, create session report | User intent (source: code/docs, optional theme) |
| 02 | run-quiz | Drive the 5-question loop: read file, audit, generate question, wait for answer, score, give feedback, adapt difficulty, update report | Session context from launch |
| 03 | flag-inconsistency | On contradiction detected between two files: notify user, create task file, log in report | Two conflicting file paths + contradiction description |
| 04 | end-session | Display final score/grade, detail per question, list weak points, finalize report, suggest replay | Completed session report |
Sequential: 01 → 02 → 04. Action 03 is triggered conditionally from within 02 whenever a contradiction is detected — it does not block the quiz loop.
launch, if a background agent is actively modifying source files, warn the user — quiz questions may be based on in-progress code.aidd_docs/tasks/<YYYY_MM>/<YYYY_MM_DD>-quiz-<N>.md after every scored question — never batch-write at the end.assets/quiz_report.md — session report template (copy to aidd_docs/tasks/<YYYY_MM>/<YYYY_MM_DD>-quiz-<N>.md at launch)CLAUDE.md — project context injected at session start; read once during launch for stack and conventions awarenessaidd_docs/memory/ — memory bank scanned when source = docs.claude/rules/ — rules cross-referenced during file audit in run-quizaidd_docs/ — user stories and decisions cross-referenced during file audit