A skill for Claude Code and Codex that conducts mock technical interviews with adaptive difficulty, structured feedback, and session history.
The best way to find out what you don't know is to have someone ask.
Interviewer turns your agent into a technical interviewer. Pick your topics, choose a difficulty mode, and answer questions one by one. Each answer gets scored on accuracy, completeness, and clarity. When you're done, you get a summary of strengths, weak areas, and what to study next. Sessions are saved so you can track progress over time and replay questions you struggled with.
Invoke with interview.
- Why
- How It Works
- Installation
- Usage
- Session Configuration
- During a Session
- After a Session
- Contributors
Reading notes and watching lectures feels like learning. But when someone asks you to explain a concept from scratch -- no hints, no looking things up -- gaps appear fast. Interviewer creates that pressure on demand:
- Active recall -- forces you to retrieve and articulate concepts, not just recognize them
- Calibrated difficulty -- adaptive mode adjusts to your level; progressive mode ramps up within a session
- Immediate feedback -- scored on three dimensions with detailed explanations after every answer
- Spaced repetition -- replay mode resurfaces questions you got wrong or skipped
┌─────────────────────────────────────────────────┐
│ 1. Setup: pick topics, format, difficulty │
├─────────────────────────────────────────────────┤
│ 2. Warm-up: 1-2 easy questions │
├─────────────────────────────────────────────────┤
│ 3. Interview loop: │
│ question → answer → score → feedback │
│ ↓ │
│ hints / skip / bookmark available │
├─────────────────────────────────────────────────┤
│ 4. Summary: scores, strengths, weak areas │
├─────────────────────────────────────────────────┤
│ 5. Save session to history/ │
└─────────────────────────────────────────────────┘
Clone into your agent's skills directory:
Claude Code:
git clone https://github.com/chrisliu298/interviewer.git ~/.claude/skills/interviewerCodex:
git clone https://github.com/chrisliu298/interviewer.git ~/.codex/skills/interviewer/interview # Start new session or list existing
/interview [session-name] # Resume or create named session
/interview replay [session] # Re-practice weak questions
/interview history # List past sessions
/interview stats # Overall progress across sessions
/interview bookmarks # Review saved favorite/tricky questions
Start a session and the agent walks you through setup, then begins asking questions. Say "done" at any time to end the session and get your summary.
Each session begins with a short setup phase where you choose:
| Option | Choices |
|---|---|
| Source | File paths, topic keywords, or both |
| Format | Standalone, multi-part with follow-ups, progressive chains |
| Types | Conceptual only, + pseudocode, full range, design-focused |
| Difficulty | Fixed (L1-L4), adaptive, progressive, random mix |
| Length | Specific count, fixed (5), or open-ended |
| Persona | Neutral/helpful, challenging/probing, Socratic |
| Scope | LLM-focused, broader ML/AI, general CS/systems |
File-based mode reads your notes or papers and generates questions from them. Topic mode generates questions from the agent's knowledge. Both mode combines the two.
Each answer is scored on three dimensions (1-5):
- Accuracy -- correctness of information
- Completeness -- coverage of key points
- Clarity -- how well explained
Mid-session commands:
| Command | Effect |
|---|---|
hint |
Get a nudge in the right direction |
skip |
Move to next question (max 2 per session) |
bookmark |
Mark question for later review |
answer |
Show the model answer |
feedback |
Get a progress check |
done |
End session and get summary |
Sessions are saved to history/ with full question-answer logs, scores, and the session summary. Use the history commands to review past performance:
/interview history-- list all past sessions with dates and scores/interview stats-- aggregate performance across sessions, improvement trends, consistent weak areas/interview replay [session]-- re-practice questions you scored below 3/5, skipped, or bookmarked/interview bookmarks-- review all bookmarked questions across sessions
- @chrisliu298
- Claude Code -- session workflow design and question template library