/onboarding

Collect identity, portfolio focus, risk thresholds, tracked segments, and location. Persist to the marketcheck-profile.md project memory file.

Step 0: Check for existing profile

Read the marketcheck-profile.md project memory file. Parse the JSON content after the --- frontmatter delimiter. If valid JSON: show summary, ask update or keep. If keep, stop.

Step 1: Collect identity

• "What is your name?" -> user.name
• "What company?" -> user.company

Use $ARGUMENTS as name or company if provided.

Step 2: Collect country

Ask: "US or UK?" Default US. Determines ZIP vs postcode, available tools (US=full, UK=limited).

Step 3: Collect location

US: ZIP (5-digit), state (2-letter), optionally city. UK: Postcode, region/county.

Step 4: Portfolio focus

Ask: "Primary business focus?" -> lender.portfolio_focus:

• auto_loans -- retail auto lending (direct/indirect)
• leasing -- vehicle leasing / residual value management
• floor_plan -- floor plan financing for dealer inventory

Step 5: LTV thresholds

Present with defaults:

• LTV warning threshold: 100% (flag when loan exceeds value)
• LTV high-risk threshold: 120%

Store as lender.risk_ltv_threshold and lender.high_risk_ltv_threshold.

Step 6: Tracked segments

Ask: "Which vehicle segments to track for risk monitoring?" Suggest: SUV, Sedan, Truck, EV, Luxury, Subcompact. Store as lender.tracked_segments.

Step 7: States of operation

Ask: "Which states do you operate in?" (comma-separated, or "national"). Store as lender.tracked_states.

Step 8: Write profile

Write to the marketcheck-profile.md project memory file with this frontmatter:

---
name: marketcheck-profile
description: Full MarketCheck user profile — identity, role, location, preferences. Read by all plugin skills and commands.
type: user
---

Then the profile JSON:

{
  "schema_version": "1.0",
  "user_type": "lender",
  "created_at": "[ISO]", "updated_at": "[ISO]",
  "user": { "name": "", "company": "" },
  "lender": {
    "portfolio_focus": "auto_loans|leasing|floor_plan",
    "risk_ltv_threshold": 100, "high_risk_ltv_threshold": 120,
    "tracked_segments": [], "tracked_states": []
  },
  "location": {
    "country": "US|UK", "zip": null, "postcode": null,
    "state": null, "region": null, "city": null
  },
  "preferences": { "default_radius_miles": 75 }
}

Set null for non-applicable location fields.

Step 9: Sync to session memory

Write a compact profile summary to your auto-memory file (MEMORY.md) so all future chat windows automatically know this user. Append or update the ## MarketCheck Profile section:

## MarketCheck Profile: Lender
- **User**: [name] | **Company**: [company] | **Focus**: [portfolio_focus]
- **Country**: [country] | **Location**: [zip/postcode], [state/region]
- **LTV thresholds**: warning=[risk_ltv]%, high-risk=[high_risk_ltv]%
- **Segments**: [tracked_segments list] | **States**: [tracked_states or "National"]
- **Profile**: marketcheck-profile.md (project memory)

Do not overwrite other memory content.

Step 10: Confirm and suggest next steps

Show profile summary: name, company, focus, LTV thresholds, tracked segments, states, location.

Next steps:

• "Depreciation rate for Tesla Model 3" -- residual value tracking
• "EV vs ICE depreciation comparison" -- portfolio risk assessment
• "Revalue these VINs [paste list]" -- portfolio spot-check
• "Collateral value for VIN [paste VIN]" -- single-unit check

UK users: UK data is active listings and recent cars only. ML pricing, sold analytics, VIN history are US-only.