Job Preferences

A Codex and Claude Code skill for managing persistent job search preferences. Set your target titles, salary floor, remote preference, and exclusion filters once; the bundled job-search skill reuses them automatically.

Workflow

Step 1: Load Profile

Follow the bundled answer-memory skill ($job-apply:answer-memory in Codex; /job-apply:answer-memory in Claude Code). Resolve <plugin-root> as that skill directs, run python3 "<plugin-root>/scripts/job-apply-store.py" init, then load saved preferences with preferences-get. Never read or write persistent Job Apply files directly.

Step 2: Check for Existing Preferences

Use the JSON object returned by preferences-get.

If preferences exist, display them:

Your current job search preferences:

- Target titles: Staff AI Engineer, Principal ML Engineer
- Min base salary: $250K
- Remote preference: Remote only
- Exclude patterns: junior, associate, intern, entry level
- Default time range: last week

Would you like to update any of these?

Then wait for the user. If they say no, stop. If they want to update, ask only about the fields they want to change using the host's structured question surface when available, or concise direct questions otherwise.

If no preferences exist, run the full Q&A below.

Step 3: Collect Preferences (full Q&A)

Use the host's structured question surface when available; otherwise ask concise direct questions. Ask no more than 4 questions at a time.

Question batch 1:

1. Target titles (multi-select + custom)

- Header: "Titles" - Question: "Which job titles are you targeting?" - Options: Staff AI Engineer, Principal ML Engineer, Director of AI, Head of ML - Multi-select: true - The user can add custom titles via "Other"

1. Min base salary

- Header: "Salary" - Question: "What is your minimum base salary?" - Options: $200K, $250K, $300K - Multi-select: false

1. Remote preference

- Header: "Remote" - Question: "What is your remote work preference?" - Options: Remote only, Remote preferred, Open to hybrid, Open to all - Multi-select: false

1. Exclude patterns (multi-select)

- Header: "Exclude" - Question: "Which patterns should be excluded from results?" - Options: junior, associate, intern, entry level - Multi-select: true

Question batch 2:

1. Default time range

- Header: "Time range" - Question: "What default time range should job searches use?" - Options: Last week, 2 weeks, Month - Multi-select: false

Step 4: Save Preferences

Write the collected values to a private temporary JSON object, then call preferences-set --input <preferences.json> through the bundled helper and remove the temporary file. The helper merges supplied keys while preserving the rest of the preferences and profile. Never patch profile.json directly.

Schema:

{
  "preferences": {
    "targetTitles": ["Staff AI Engineer", "Principal ML Engineer"],
    "minBaseSalary": "$250K",
    "remotePreference": "remote only",
    "excludePatterns": ["junior", "associate", "intern", "entry level"],
    "defaultTimeRange": "last week"
  }
}

Step 5: Confirm

Display the saved preferences and confirm:

Preferences saved to your local Job Apply store at ~/.job-apply/profile.json.

These will be used automatically by the bundled job-search skill. Invoke job-preferences again any time to update them.

Updating Preferences

When the user invokes job-preferences and preferences already exist, show current values and let them update selectively. Only overwrite the fields they change — keep the rest intact.

Safety Rules

1. Use only the helper — initialize, read, and merge preferences through the bundled answer-memory skill; never directly edit files under ~/.job-apply/
2. Preserve existing profile — use preferences-set without --replace for selective updates
3. No defaults without user input — always ask the user, never assume values