How the `/interview` Command Builds Stage-Specific Interview Preparation Packs

The /interview command follows a deterministic, five-step pipeline that transforms tracked application data into stage-specific interview preparation packs by combining archived submission materials, verified company research, and reusable framework assets.

The MadsLorentzen/ai-job-search repository provides a Claude-powered job application system where the /interview command automatically generates stage-specific interview preparation packs. This deterministic pipeline ensures every talking point aligns with your submitted CV and cover letter while incorporating independently verified company intelligence tailored to the specific interview round.

The Five-Step Pipeline for Stage-Specific Generation

Step 0: Parse Input and Locate Target Applications

In .claude/commands/interview.md, the command first receives an optional argument (e.g., /interview acme). It queries job_search_tracker.csv to locate the company and role. When no argument is provided, the system lists all rows with statuses of interview, offer, or recently applied, allowing you to select from active candidacies.

Step 1: Load Application Context from Archives

The command retrieves the Application Archive created by the /outcome command. It specifically reads:

• documents/applications/<company>_<role>/job_posting.md
• cv_draft.tex and cover_letter.tex (the exact files the recruiter reviewed)
• outcome.md (containing stage history and prior feedback)

If any archive components are missing, the system falls back to generic CV/cover-letter files while explicitly informing you which specific pieces are absent.

Step 2: Execute Interview-Focused Company Research

Following the Company Research Checklist defined in .claude/skills/job-application-assistant/04-job-evaluation.md, the command gathers mission statements, recent news, LinkedIn data, and media coverage via WebFetch. When interviewer names are known, it also collects their public profiles to infer likely interview angles. Every fact is verified before inclusion to ensure accuracy.

Step 3: Generate the Stage-Specific Prep Pack

The system creates documents/applications/<company>_<role>/interview_prep_<stage>.md—one file per interview stage (e.g., phone screen, technical, final). Each pack contains six critical sections:

1. Likely questions – prioritized from outcome feedback, fit-evaluation gaps, posting requirements, and stage type.
2. STAR answer mapping – reusing existing STAR examples from 07-interview-prep.md or drafting new ones strictly from 01-candidate-profile.md.
3. Consistency brief – a defensive checklist of claims made in your CV and cover letter that must be verifiable.
4. Tough, customized questions – drawn from 07-interview-prep.md with verified company-specific hooks.
5. Questions to ask – curated from the same source and tuned to the interview stage.
6. Logistics – format-specific tips, interview date, and interviewer names.

Step 4: Offer Interactive Mock Interviews

The command prompts you to role-play the interview. If accepted, the dialog follows the "Roleplay Guidelines" inside .claude/skills/job-application-assistant/07-interview-prep.md, delivering feedback calibrated against .claude/skills/job-application-assistant/02-behavioral-profile.md. This step produces no file output—it operates as an interactive chat flow.

Step 5: Close the Loop with Outcome Tracking

Finally, the command reminds you to execute /outcome <company> after the interview to store new feedback for subsequent rounds. If you approved any new STAR answers during preparation, it confirms these have been appended to 07-interview-prep.md for future use.

Core Architecture Behind Stage-Aware Generation

Three architectural concepts enable this stage-specific generation:

Application Archives – Per-company/role folders under documents/applications/ store the exact job posting, submitted CV/cover letter, and ongoing outcome logs, ensuring the prep pack reflects your actual submission materials.

Framework Assets – Reusable knowledge files including 07-interview-prep.md (STAR templates and tough questions), 04-job-evaluation.md (research criteria), and 01-candidate-profile.md (candidate facts) provide vetted interview strategies and behavioral benchmarks.

Stage-Aware Logic – Each interview stage (phone screen, technical, final) influences question prioritization, logistics tone, and the specific set of "questions to ask," guaranteeing the output matches the round's unique demands.

Generating and Accessing Your Prep Pack

Trigger the pipeline using the command interface:

# Example: invoking the command from a Claude-compatible CLI

import subprocess

# Prepare interview prep for the Acme Inc. Senior Engineer role

company = "Acme"
role = "Senior Engineer"
subprocess.run(["claude", "interview", f"{company}_{role}"])

# Output location:

# documents/applications/Acme_Senior Engineer/interview_prep_technical.md

Inspect the generated pack directly:

# View the stage-specific preparation file

$ cat documents/applications/Acme_Senior\ Engineer/interview_prep_technical.md

# Displays: Likely questions, STAR mappings, Consistency brief, etc.

You can also preview content without writing files by asking the assistant:

/interview Acme Senior Engineer

The assistant displays the markdown content while simultaneously writing it to the archive location.

Summary

• The /interview command operates via a deterministic five-step pipeline defined in .claude/commands/interview.md.
• It pulls from Application Archives (documents/applications/<company>_<role>/) containing your exact submission materials and history.
• Company Research Checklist execution in Step 2 ensures all facts are verified before inclusion.
• Output files follow the naming pattern interview_prep_<stage>.md and contain six structured sections including STAR mappings and consistency briefs.
• New STAR answers approved during the process are automatically appended to 07-interview-prep.md.

Frequently Asked Questions

What files does the /interview command read to build preparation packs?

The command reads the Application Archive created by /outcome, specifically job_posting.md, cv_draft.tex, cover_letter.tex, and outcome.md from documents/applications/<company>_<role>/. It also references framework assets like 07-interview-prep.md and 01-candidate-profile.md for STAR templates and candidate facts.

How does the system ensure the preparation pack matches the specific interview stage?

Stage-Aware Logic adjusts the priority of likely questions, the tone of logistics sections, and the curated "questions to ask" based on whether you are facing a phone screen, technical interview, or final round. Each stage generates a separate interview_prep_<stage>.md file.

Where are the generated stage-specific interview preparation packs stored?

Files are written to documents/applications/<company>_<role>/interview_prep_<stage>.md within the repository's application tracking structure. This location is git-ignored by default to protect sensitive interview details.

Can I practice with the generated content before the actual interview?

Yes. Step 4 offers an interactive mock interview that follows the "Roleplay Guidelines" in 07-interview-prep.md and provides feedback calibrated against your 02-behavioral-profile.md. This simulation uses the content from your prep pack without creating additional files.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}