AgentSkillsCN

mock

模拟一轮完整的面试过程，配备计时问题、符合人物设定的后续追问，以及回合结束时的综合评估。

SKILL.md
--- frontmatter
name: mock
description: "Simulate a full interview round with timed questions, persona-consistent follow-ups, and end-of-round assessment"
disable-model-invocation: true
allowed-tools: Read, Write, Edit, Bash, Glob, Grep, AskUserQuestion
argument-hint: "[company: amazon|google|meta|<custom>] [--stage phone-screen|technical|onsite|bar-raiser|system-design] [--duration 30|45|60]"

Mock Interview

Run a realistic interview round: multiple questions, consistent persona, timed pacing, and a comprehensive debrief at the end.

Steps

1. Load Configuration

Read .claude/what-did-you-do.local.md and .claude/i-did-a-thing.local.md. If either is missing, direct the user to the appropriate setup.

2. Load Arsenal and Session History

Read <things_path>/arsenal/ and <things_path>/targets/profile.md. Run bash <plugin_root>/scripts/search-sessions.sh --type mock --recent 5 to load recent mock history.

3. Select Company and Stage

If the user provided a company argument, load the company profile from <plugin_root>/companies/<company>.yaml or <things_path>/interview-prep/companies/<company>.yaml for custom profiles.

If no company specified, use AskUserQuestion:

Which company are you preparing for?

  • •Amazon — LP-driven, bar raiser focus
  • •Google — technical excellence, Googleyness
  • •Meta — move fast, impact-driven
  • •Generic — no company-specific framing
  • •Custom — I have a company profile set up

If the user provided a --stage, use it. Otherwise, if a company was selected, present the company's interview stages. If generic, use AskUserQuestion:

Which interview stage?

  • •Phone Screen (30 min) — 3-4 questions, culture and motivation
  • •Technical (45 min) — coding and technical discussion
  • •Onsite Behavioral (45 min) — 4-5 deep behavioral questions
  • •System Design (45 min) — one end-to-end design problem
  • •Bar Raiser (60 min) — cross-functional deep dive

4. Configure the Round

Based on company profile and stage, determine:

  • •Duration: from --duration arg, company profile, or stage default
  • •Number of questions: from references/mock-formats.md
  • •Persona: from company profile's persona field for this stage, or select appropriate persona
  • •Evaluation focus: from company profile or stage defaults
  • •Question categories: from company values → question_themes mapping

Load the selected persona from <plugin_root>/personas/<persona>.md.

5. Brief the User

Mock Interview: [Company] — [Stage]

Duration: ~[X] minutes Questions: [N] Interviewer: [Persona Role] Focus: [evaluation areas]

I'll ask questions one at a time. Answer as you would in a real interview. I'll hold all feedback until the debrief at the end.

Ready?

Wait for confirmation.

6. Conduct the Round

For each question in the round:

  1. •Select a question using the same spaced repetition algorithm as the practice skill, but filtered by the company's question_themes and stage's evaluation_focus
  2. •Ask the question in the persona's voice
  3. •Receive the answer
  4. •If the question has relevant depth: 2 follow-ups and time permits, ask ONE follow-up in persona voice
  5. •Silently score the answer (do NOT share scores yet)
  6. •Track time spent per question (use references/timing-guide.md)
  7. •Transition to next question naturally in persona voice

Between questions, use persona-appropriate transitions:

  • •Staff Engineer: "Good. Let me shift gears..."
  • •Engineering Manager: "Thank you for sharing that. I'd like to explore..."
  • •Bar Raiser: "Interesting. Here's a different angle..."

Do NOT provide feedback between questions. Save it all for the debrief.

7. End-of-Round Debrief

After all questions, break character and deliver a comprehensive assessment:

Mock Interview Debrief

Overall Assessment: [Strong / Advancing / Developing / Not Ready]

Dimension Scores:

DimensionAvg ScoreStrongest AnswerWeakest Answer
Specificityx/5Q<n>Q<n>
Structurex/5Q<n>Q<n>
Impactx/5Q<n>Q<n>
Relevancex/5Q<n>Q<n>
Self-Advocacyx/5Q<n>Q<n>

If company-specific, add:

[Company] Alignment:

  • •Values demonstrated: <list>
  • •Values missing: <list>
  • •Level calibration: <assessment vs target level>

Then per-question breakdown:

Q1: "<question>" Score: x/5 | Strengths: <brief> | To improve: <brief> Arsenal suggestion: <relevant log> (<evidence_type>) — e.g., "Your lesson about X is directly relevant" or "Your decision entry on Y demonstrates the tradeoff analysis they want"

Finally:

Top 3 Action Items:

  1. • <most impactful improvement>
  2. • <next priority>
  3. • <third priority>

8. Log the Session

Write session file at <things_path>/interview-prep/sessions/<date>-mock-<company>-<stage>.md with full frontmatter including all per-question scores, company alignment, and overall assessment.

9. Update Progress Dashboard

Update <things_path>/interview-prep/progress.md with new scores and session entry.

10. Suggest Unlogged Accomplishments

If the user mentioned experiences not in their arsenal during the mock:

During the mock, you referenced some experiences I don't see in your logs:

  • •<experience> — this sounds like a <evidence type> entry

Logging these would strengthen your arsenal. Run /i-did-a-thing:thing-i-did when ready.

11. Offer Next Steps

Use AskUserQuestion:

What next?

  • •Run another mock round (same company, different stage)
  • •Drill specific questions that were weak
  • •Review progress across all mocks
  • •Prep for a specific company (/what-did-you-do:prep-for)
  • •Done for now