Transactionsyncing

TransactionSyncing

Import Fidelity transaction history CSV into Google Sheets using a hybrid architecture: master Transactions tab for full audit trail + auto-routing of debit card purchases to Expense Tracker for Budget Planner integration.

Workflow Routing

When executing this workflow, output this notification:

Running the **SyncTransactions** workflow from the **TransactionSyncing** skill...

Typical flow: IngestTransactions (local archive) -> SyncTransactions (Google Sheets)

Examples

Example 1: Sync after downloading Fidelity transaction history

User: "sync transactions"
-> Reads History_for_Account_{account_id}.csv from notebooks/transactions/
-> Creates/updates Transactions tab with full Fidelity data
-> Routes DEBIT CARD PURCHASE entries to Expense Tracker
-> Auto-categorizes expenses (H-E-B -> Groceries, Tesla -> Auto & Transport)
-> Reports: "Added 45 transactions, 12 expenses categorized"

Example 2: Import new transaction export

User: "import the transaction history"
-> Invokes SyncTransactions workflow
-> Detects duplicates by date + action + amount
-> Skips existing entries, adds only new ones
-> Flags uncategorized expenses for manual review

Example 3: Check recent transactions

User: "import fidelity transactions and update expense tracker"
-> Full sync with expense routing
-> Generates summary of dividends received, purchases, margin interest

Architecture Overview

Data Flow

Fidelity CSV (notebooks/transactions/)
         |
         v
+-------------------+
| Transactions Tab  |  <- Master source (ALL transactions)
| (Full Fidelity)   |
+-------------------+
         |
         | Filter: DEBIT CARD PURCHASE
         v
+-------------------+
| Expense Tracker   |  <- Budget Planner integration
| (Categorized)     |
+-------------------+

Transaction Types Handled

Smart Categorization

See CategoryRules.md for the full pattern matching rules.

Sample patterns:

• H-E-B, KROGER, COSTCO, WAL-MART -> Groceries
• Tesla, SUPERCHA -> Auto & Transport
• BENIHANA, GOLDEN CORRAL, PAPA JOHN -> Dining Out
• CVS, PHARMACY -> Health & Wellness

Core Workflow

1. Read Fidelity Transaction History CSV

Location: notebooks/transactions/History_for_Account_{account_id}.csv

CSV Columns:

Run Date, Action, Symbol, Description, Type, Price ($), Quantity,
Commission ($), Fees ($), Accrued Interest ($), Amount ($),
Cash Balance ($), Settlement Date

2. Create/Update Transactions Tab

Google Sheets Structure:

3. Deduplicate

Match criteria: Date + Action + Amount

For each CSV row:
  key = f"{run_date}|{action}|{amount}"
  if key exists in sheet:
    SKIP (already imported)
  else:
    ADD to Transactions tab

4. Route Expenses to Expense Tracker

Filter: Action contains "DEBIT CARD PURCHASE"

Expense Tracker Format:

Category Assignment: See CategoryRules.md

5. Generate Summary

SYNC SUMMARY - [Date]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

TRANSACTIONS TAB:
  New entries: 45
  Skipped (duplicates): 12

EXPENSE TRACKER:
  Expenses routed: 18
  Auto-categorized: 15
  Needs review: 3

BY TYPE:
  Dividends: $342.50
  Margin Interest: -$18.43
  Debit Card: -$1,245.67
  Direct Deposit: +$5,054.09
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Google Sheets Integration

Spreadsheet ID: Read from fin-guru/data/user-profile.yaml -> google_sheets.portfolio_tracker.spreadsheet_id

Creating Transactions Tab (if needed)

// Check if Transactions tab exists
mcp__gdrive__sheets(operation: "listSheets", params: {
    spreadsheetId: SPREADSHEET_ID
})

// Create if missing
mcp__gdrive__sheets(operation: "createSheet", params: {
    spreadsheetId: SPREADSHEET_ID,
    title: "Transactions"
})

// Add headers
mcp__gdrive__sheets(operation: "updateCells", params: {
    spreadsheetId: SPREADSHEET_ID,
    range: "Transactions!A1:I1",
    values: [["Date", "Action", "Symbol", "Description", "Type", "Amount", "Category", "Balance", "Settlement"]]
})

Adding to Expense Tracker

// Append expense row
mcp__gdrive__sheets(operation: "appendRows", params: {
    spreadsheetId: SPREADSHEET_ID,
    sheetName: "Expense Tracker",
    values: [[date, description, category, amount, month]]
})

Critical Rules

WRITABLE Destinations

• Transactions tab: All columns (new tab, we control format)
• Expense Tracker: Append new rows only (preserve existing)

NEVER MODIFY

• Budget Planner formulas
• Existing Expense Tracker entries

Deduplication Key

• Transactions tab: Date|Action|Amount
• Expense Tracker: Date|Description|Amount

Reference Files

• CategoryRules.md: Pattern matching rules for expense categorization
• fin-guru/data/user-profile.yaml: Spreadsheet ID
• scripts/google-sheets/portfolio-optimizer/: Apps Script reference

Pre-Flight Checklist

Before syncing transactions:

• [ ] Transaction History CSV exists in notebooks/transactions/
• [ ] CSV is from Fidelity (not other broker)
• [ ] Expense Tracker tab exists in Google Sheets
• [ ] Current date retrieved via date command

---

Skill Type: Domain (workflow guidance) Enforcement: SUGGEST Priority: Medium Line Count: < 300 (following 500-line rule)

CategoryRules - Expense Categorization Patterns

Pattern matching rules for auto-categorizing debit card purchases from Fidelity transaction history.

How to Add New Patterns

To add a new categorization rule:

1. Identify the merchant name pattern from Fidelity descriptions 2. Add to the appropriate category section below 3. Patterns are case-insensitive 4. Use partial matches (e.g., "h-e-b" matches "H-E-B #063 Pearland TX")

---

Category Patterns

Groceries

Supermarkets, grocery stores, food supplies

Patterns:

• h-e-b, heb
• kroger
• costco
• wal-mart, walmart
• wholefds, whole foods
• makola
• target (when food context)
• sam's club
• aldi
• trader joe

Examples:

• "H-E-B #063 Pearland TX" -> Groceries
• "COSTCO WHSE #1 PEARLAND TX" -> Groceries
• "MAKOLA IMPORTS HOUSTON TX" -> Groceries

---

Dining Out

Restaurants, fast food, entertainment dining

Patterns:

• benihana
• golden corral
• papa john
• chuck e cheese
• wingstop
• cinemark
• mcdonald
• chick-fil-a
• chipotle
• starbucks
• coffee
• restaurant
• grill
• cafe
• makiin
• sparkly photo (event dining)

Examples:

• "BENIHANA SUGAR LAND" -> Dining Out
• "TST*MAKIIN Houston TX" -> Dining Out
• "PAPA JOHN'S #2 PEARLAND TX" -> Dining Out

---

Auto & Transport

Vehicle expenses, fuel, parking, transportation

Patterns:

• tesla
• supercha (Tesla Supercharger)
• parking
• fastpark
• uber
• lyft
• shell
• exxon
• chevron
• valero
• buc-ee
• gas station
• toll

Examples:

• "Tesla, Inc. SUPERCHA600118984238637" -> Auto & Transport
• "FASTPARKHOU HOUSTON TX" -> Auto & Transport
• "Tesla Property Casual Fremont CA" -> Auto & Transport

---

Personal Care

Grooming, beauty, self-care

Patterns:

• salon
• spa
• barber
• sephora
• beauty supply
• supreme beauty
• ulta
• nail
• hair
• shaving grace
• gloss* skin
• cash app* (often personal transfers)

Examples:

• "K STAR SALON & SPA MANVEL TX" -> Personal Care
• "A SHAVING GRACE BARBER PEARLAND TX" -> Personal Care
• "SEPHORA PEACHT PEACHTREE CI GA" -> Personal Care

---

Health & Wellness

Medical, pharmacy, fitness

Patterns:

• cvs
• pharmacy
• walgreens
• life time (gym)
• doctor
• medical
• dental
• clinic
• hospital
• urgent care

Examples:

• "CVS/PHARMACY # MANVEL TX" -> Health & Wellness
• "LIFE TIME #320" -> Health & Wellness

---

Shopping

Retail, clothing, general merchandise

Patterns:

• marshalls
• target (non-food)
• amazon
• skims
• tj maxx
• ross
• old navy
• gap
• nordstrom
• macy
• best buy
• apple store

Examples:

• "MARSHALLS #877 PEARLAND TX" -> Shopping
• "SP SKIMS CHECKOUT.SKIM CA" -> Shopping

---

Family Care

Childcare, family activities, kids

Patterns:

• aqua tots
• brightwheel, brghtwhl
• daycare
• childcare
• school
• kid
• children
• pediatric

Examples:

• "AQUA TOTS - PEARLAND" -> Family Care
• "BRGHTWHL R* REDEEMER" -> Family Care

---

Bills & Utilities

Recurring bills, subscriptions, utilities

Patterns:

• autopay
• acctverify
• electric
• water
• internet
• comcast
• att
• verizon
• t-mobile
• netflix
• spotify
• subscription

Examples:

• "BMO ACCTVERIFY" -> Bills & Utilities

---

Cash Withdrawal

ATM and cash transactions

Patterns:

• atm
• cash withdrawal
• cash advance

Examples:

• "ATM0043 11555 MAGNOLIA PEARLAND TX" -> Cash Withdrawal
• "ATMXD10 *SEDONA LAKES MANVEL TX" -> Cash Withdrawal

---

Tuition

Education expenses

Patterns:

• regent univer
• university
• college
• tuition
• school
• education
• coursera
• udemy

Examples:

• "REGENT UNIVERSPURCHASE" -> Tuition

---

Business Expense

Work-related purchases

Patterns:

• gumroad
• ups
• fedex
• office depot
• staples
• postal
• usps
• business
• linkedin
• zoom

Examples:

• "GUMROAD* SHAWN GRADY" -> Business Expense
• "POSTAL COPY CENTER-931 PEARLAND TX" -> Business Expense

---

Loan Payment

Debt payments

Patterns:

• wells fargo + draft or audraft
• loan payment
• mortgage
• car payment
• student loan
• credit card payment

Examples:

• "WELLS FARGO AUDRAFT" -> Loan Payment

---

Home & Garden

Home improvement, garden, maintenance

Patterns:

• home depot
• lowes
• sawyer
• smart core
• garden
• hardware
• furniture

Examples:

• "SAWYER + S* SMART CORE" -> Home & Garden

---

Crypto Deposit

Cryptocurrency deposits and transfers

Patterns:

• btc deposited
• bitcoin
• fidelity crypto
• eth deposited
• crypto

Examples:

• "0.17713256 BTC deposited" -> Crypto Deposit
• "Fidelity Crypto® 8449251033" -> Crypto Deposit

---

Credit Card Payment

Credit card bill payments

Patterns:

• applecard
• gsbapayment
• chase payment
• amex payment
• discover payment

Examples:

• "DIRECT DEBIT APPLECARD GSBAPAYMENT" -> Credit Card Payment

---

Exempt

Verification transactions, zero amounts

Patterns:

• ifacctverify
• verification
• Amount = $0.00 or < $1.00

Examples:

• "WELLS FARGO IFACCTVERIFY" -> Exempt

---

Uncategorized

Any transaction not matching the above patterns is marked as "Uncategorized" and flagged for manual review in the sync summary.

Common uncategorized reasons:

• New merchant not in patterns
• Unusual description format
• One-time or rare purchase

To resolve: Add the pattern to the appropriate category above.

---

Pattern Matching Algorithm

def categorize_expense(description: str) -> str:
    desc = description.lower()

    # Check each category's patterns
    for category, patterns in CATEGORY_PATTERNS.items():
        for pattern in patterns:
            if pattern in desc:
                return category

    return "Uncategorized"

CATEGORY_PATTERNS = {
    "Groceries": ["h-e-b", "heb", "kroger", "costco", "wal-mart", "walmart", ...],
    "Dining Out": ["benihana", "golden corral", "papa john", ...],
    "Auto & Transport": ["tesla", "supercha", "parking", "fastpark", ...],
    # ... etc
}

---

Extending Categories

When the user wants to add new categories or patterns:

1. Add to existing category: Update the patterns list above 2. Create new category: Add a new section with patterns and examples 3. Expense Tracker sync: Ensure the category name matches Budget Planner

Budget Planner categories (must match exactly):

• Groceries
• Dining Out
• Auto & Transport
• Personal Care
• Health & Wellness
• Shopping
• Family Care
• Bills & Utilities
• Cash Withdrawal
• Tuition
• Business Expense
• Loan Payment
• Home & Garden
• Crypto Deposit
• Credit Card Payment
• Exempt
• Software & Tech
• Cell Phone
• Gas
• Water
• Light Bill
• Mortgage

---

Last Updated: 2026-01-02 Maintainer: Finance Guru TransactionSyncing skill

IngestTransactions Workflow

Ingest Fidelity transaction history CSV from Downloads into the local rolling archive. Detects date range (30d/60d), creates date-stamped copy, merges unique rows into Accounts_History.csv.

Triggers

• "ingest transactions", "import history", "bring in transactions"
• User points to ~/Downloads/History_for_Account_Z05724592.csv
• User mentions downloading transaction history from Fidelity

Step 1: Locate Source File

Default location: ~/Downloads/History_for_Account_Z05724592.csv Alternative: User may specify a different path.

# Verify file exists
ls -la ~/Downloads/History_for_Account_Z05724592.csv

If not found, ask user for the file location.

Step 2: Read and Analyze the CSV

CSV Schema (13 columns, header on row 3, rows 1-2 are blank):

Run Date, Action, Symbol, Description, Type, Price ($), Quantity,
Commission ($), Fees ($), Accrued Interest ($), Amount ($),
Cash Balance ($), Settlement Date

Detect date range:

# Extract earliest and latest dates from the CSV (skip header rows, skip footer disclaimer)
awk -F',' 'NR>2 && $1 ~ /^[0-9]/ {print $1}' "$SOURCE_FILE" | sort -t/ -k3,3n -k1,1n -k2,2n | head -1  # earliest
awk -F',' 'NR>2 && $1 ~ /^[0-9]/ {print $1}' "$SOURCE_FILE" | sort -t/ -k3,3n -k1,1n -k2,2n | tail -1  # latest

Classify period:

• Calculate days between earliest and latest date
• <= 35 days = 30d
• <= 65 days = 60d
• > 65 days = {N}d (use actual count)

Count transaction rows (exclude blank lines and Fidelity disclaimer footer):

awk -F',' 'NR>2 && $1 ~ /^[0-9]/ {count++} END {print count}' "$SOURCE_FILE"

Step 3: Copy with Date-Stamped Name

Destination: notebooks/transactions/

Naming convention: History_for_Account_Z05724592_{YYYY-MM-DD}_{period}.csv

Where:

• {YYYY-MM-DD} = the download/run date (latest date in file, or today's date)
• {period} = 30d, 60d, or {N}d

Example: History_for_Account_Z05724592_2026-03-06_60d.csv

cp ~/Downloads/History_for_Account_Z05724592.csv \
   notebooks/transactions/History_for_Account_Z05724592_2026-03-06_60d.csv

Step 4: Merge into Accounts_History.csv

Master archive: notebooks/transactions/Accounts_History.csv

Schema Normalization

The master archive may have a 14-column legacy schema (with Account, Account Number columns). New Fidelity downloads use a 13-column schema. Normalize as follows:

Canonical 13-column schema (used going forward):

Run Date, Action, Symbol, Description, Type, Price ($), Quantity,
Commission ($), Fees ($), Accrued Interest ($), Amount ($),
Cash Balance ($), Settlement Date

If the existing Accounts_History.csv has 14 columns (Account, Account Number after Run Date), strip those columns during merge. New rows always use 13-column format.

Deduplication

Match on: Run Date + Action + Amount($)

# Pseudocode for merge logic
existing_keys = set()
for row in accounts_history:
    key = f"{row['Run Date']}|{row['Action'][:60]}|{row['Amount ($)']}"
    existing_keys.add(key)

new_rows = []
for row in new_csv:
    key = f"{row['Run Date']}|{row['Action'][:60]}|{row['Amount ($)']}"
    if key not in existing_keys:
        new_rows.append(row)

Merge Strategy

1. Read all data rows from Accounts_History.csv (skip header, skip footer disclaimer) 2. Read all data rows from new CSV (skip blank rows 1-2, skip header row 3, skip footer) 3. Deduplicate using key above 4. Combine: existing rows + new unique rows 5. Sort by Run Date descending (newest first) 6. Write back with single header row + combined data (NO footer disclaimer)

Handle Missing Archive

If Accounts_History.csv doesn't exist:

• Copy the new CSV as the initial archive
• Strip blank rows and footer disclaimer
• Add proper header row

Step 5: Update notebooks/updates/ Copy

After merging, also update notebooks/updates/History_for_Account_Z05724592.csv with the latest download so other skills (dividend-tracking, etc.) can reference it:

cp ~/Downloads/History_for_Account_Z05724592.csv \
   notebooks/updates/History_for_Account_Z05724592.csv

Step 6: Extract Dividend Summary

From the newly ingested transactions, extract all DIVIDEND RECEIVED entries and report:

DIVIDEND INCOME (from this import)
Date       | Symbol | Amount  | Type
03/06/2026 | AMZY   | $3.61   | Cash
03/06/2026 | AMZY   | $14.82  | Margin
...
TOTAL: $XXX.XX

This data supplements (not replaces) the dividends.csv forward-looking projections.

Step 7: Generate Ingestion Report

TRANSACTION INGESTION COMPLETE - {date}
---

SOURCE: ~/Downloads/History_for_Account_Z05724592.csv
PERIOD: {earliest_date} to {latest_date} ({N}d)
ARCHIVED AS: History_for_Account_Z05724592_{date}_{period}.csv

MERGE RESULTS:
  New rows added to Accounts_History: XX
  Duplicates skipped: XX
  Archive now covers: {earliest_archive_date} to {latest_archive_date}
  Total archive rows: XX

TRANSACTION BREAKDOWN:
  Dividends received:  $XXX.XX (XX entries)
  Reinvestments:       $XXX.XX (XX entries)
  Direct deposits:     +$X,XXX.XX (XX entries)
  Debit card:          -$X,XXX.XX (XX entries)
  Margin interest:     -$XX.XX (XX entries)
  Other:               $XXX.XX (XX entries)

NEXT STEPS:
  -> Run "sync transactions" to push to Google Sheets
  -> Run "sync dividends" to update Dividend Tracker
---

Gap Detection

After merge, check for date gaps > 3 business days in the archive. Report any gaps:

GAP ALERT: No transactions between {date1} and {date2} ({N} business days)
This may indicate a missing export period. Consider downloading that range.

Error Handling

File not found

Source file not found at ~/Downloads/History_for_Account_Z05724592.csv
Please download your transaction history from Fidelity (last 30 or 60 days).

Schema mismatch

WARNING: CSV schema doesn't match expected Fidelity format.
Expected 13 columns: Run Date, Action, Symbol, ...
Found: {N} columns
Please verify this is a Fidelity History export.

Duplicate archive name

If History_for_Account_Z05724592_{date}_{period}.csv already exists:

• Compare file sizes. If identical, skip copy.
• If different, append a counter: ..._2026-03-06_60d_2.csv

---

Workflow Type: Local file management Estimated Duration: 10-20 seconds Dependencies: CSV file access, notebooks/transactions/ directory Chains to: SyncTransactions (optional, for Google Sheets push)

SyncTransactions Workflow

Import Fidelity transaction history CSV into Google Sheets with smart routing and categorization.

Step 1: Read Transaction History CSV

Location: notebooks/transactions/History_for_Account_{account_id}.csv

Read the CSV and parse:

# Expected columns (Row 3 has headers - skip rows 1-2)
Run Date, Action, Symbol, Description, Type, Price ($), Quantity,
Commission ($), Fees ($), Accrued Interest ($), Amount ($),
Cash Balance ($), Settlement Date

Key Fields to Extract:

• Run Date: Transaction date (MM/DD/YYYY format)
• Action: Transaction type (DIVIDEND RECEIVED, DEBIT CARD PURCHASE, etc.)
• Symbol: Ticker symbol (if applicable)
• Description: Full description text
• Type: Cash or Margin
• Amount ($): Dollar amount (positive = credit, negative = debit)
• Cash Balance ($): Running balance after transaction

Step 2: Check/Create Transactions Tab

Read existing sheets:

mcp__gdrive__sheets(operation: "listSheets", params: {
    spreadsheetId: "{spreadsheet_id}"
})

If "Transactions" tab doesn't exist, create it:

mcp__gdrive__sheets(operation: "createSheet", params: {
    spreadsheetId: "{spreadsheet_id}",
    title: "Transactions"
})

// Add headers
mcp__gdrive__sheets(operation: "updateCells", params: {
    spreadsheetId: "{spreadsheet_id}",
    range: "Transactions!A1:I1",
    values: [["Date", "Action", "Symbol", "Description", "Type", "Amount", "Category", "Balance", "Settlement"]]
})

Step 3: Read Existing Transactions (for deduplication)

mcp__gdrive__sheets(operation: "readSheet", params: {
    spreadsheetId: "{spreadsheet_id}",
    range: "Transactions!A:F"
})

Build deduplication set:

existing_keys = set()
for row in sheet_data:
    key = f"{row['Date']}|{row['Action']}|{row['Amount']}"
    existing_keys.add(key)

Step 4: Process CSV Transactions

For each transaction in the CSV:

4a. Generate Deduplication Key

key = f"{run_date}|{action}|{amount}"
if key in existing_keys:
    continue  # Skip duplicate

4b. Assign Category

Category Assignment Rules (see CategoryRules.md for full list):

def assign_category(action, description):
    action_lower = action.lower()
    desc_lower = description.lower()

    # Investment categories (Transactions tab only)
    if "dividend" in action_lower:
        return "DIVIDEND"
    if "reinvestment" in action_lower:
        return "REINVESTMENT"
    if "margin interest" in action_lower:
        return "MARGIN_INTEREST"
    if "cap gain" in action_lower:
        return "CAP_GAIN"
    if "direct deposit" in action_lower:
        return "INCOME"
    if "journaled" in action_lower:
        return "INTERNAL_TRANSFER"

    # Expense categories (route to Expense Tracker)
    if "debit card" in action_lower:
        return categorize_expense(desc_lower)

    return "OTHER"

def categorize_expense(description):
    # Groceries
    if any(x in description for x in ['h-e-b', 'heb', 'kroger', 'costco',
            'wal-mart', 'walmart', 'wholefds', 'whole foods', 'makola']):
        return "Groceries"

    # Dining Out
    if any(x in description for x in ['benihana', 'golden corral', 'papa john',
            'chuck e cheese', 'wingstop', 'cinemark']):
        return "Dining Out"

    # Auto & Transport
    if any(x in description for x in ['tesla', 'supercha', 'parking',
            'fastpark']):
        return "Auto & Transport"

    # Personal Care
    if any(x in description for x in ['salon', 'spa', 'barber', 'sephora',
            'beauty supply', 'cash app']):
        return "Personal Care"

    # Health & Wellness
    if any(x in description for x in ['cvs', 'pharmacy', 'walgreens']):
        return "Health & Wellness"

    # Shopping
    if any(x in description for x in ['marshalls', 'target', 'amazon',
            'skims']):
        return "Shopping"

    # Family Care
    if any(x in description for x in ['aqua tots', 'brightwheel']):
        return "Family Care"

    # Bills & Utilities
    if any(x in description for x in ['autopay', 'acctverify']):
        return "Bills & Utilities"

    # Cash Withdrawal
    if 'atm' in description:
        return "Cash Withdrawal"

    # Tuition
    if 'regent univer' in description:
        return "Tuition"

    # Business Expense
    if any(x in description for x in ['gumroad', 'ups']):
        return "Business Expense"

    # Loan Payment
    if 'wells fargo' in description and 'draft' in description:
        return "Loan Payment"

    return "Uncategorized"  # Flag for manual review

4c. Build Transaction Row

transaction_row = [
    run_date,           # Date
    clean_action(action), # Action (simplified)
    symbol or "",       # Symbol
    description[:50],   # Description (truncated)
    tx_type,            # Type (Cash/Margin)
    amount,             # Amount
    category,           # Category
    balance,            # Balance
    settlement_date     # Settlement
]

Step 5: Batch Update Transactions Tab

// Append all new transactions at once
mcp__gdrive__sheets(operation: "appendRows", params: {
    spreadsheetId: "{spreadsheet_id}",
    sheetName: "Transactions",
    values: new_transaction_rows
})

Step 6: Route Expenses to Expense Tracker

Filter debit card purchases:

expense_rows = []
for tx in new_transactions:
    if "debit card" in tx['action'].lower():
        # Format for Expense Tracker
        month = get_month_name(tx['date'])  # "January", "February", etc.
        expense_rows.append([
            tx['date'],
            tx['description'],
            tx['category'],
            format_amount(tx['amount']),  # "$XX.XX" format
            month
        ])

Read existing Expense Tracker for deduplication:

mcp__gdrive__sheets(operation: "readSheet", params: {
    spreadsheetId: "{spreadsheet_id}",
    range: "Expense Tracker!A:D"
})

Append only new expenses:

mcp__gdrive__sheets(operation: "appendRows", params: {
    spreadsheetId: "{spreadsheet_id}",
    sheetName: "Expense Tracker",
    values: new_expense_rows
})

Step 7: Generate Summary Report

TRANSACTION SYNC COMPLETE - [Date]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

TRANSACTIONS TAB:
  New entries added: XX
  Duplicates skipped: XX

EXPENSE TRACKER:
  Expenses routed: XX
  Auto-categorized: XX
  Needs review (Uncategorized): XX

TRANSACTION BREAKDOWN:
  Dividends:        $XXX.XX (XX entries)
  Reinvestments:    $XXX.XX (XX entries)
  Margin Interest:  -$XX.XX (XX entries)
  Debit Card:       -$X,XXX.XX (XX entries)
  Direct Deposits:  +$X,XXX.XX (XX entries)
  Other:            $XXX.XX (XX entries)

UNCATEGORIZED EXPENSES (needs review):
  - [Date] [Description] [$Amount]
  - [Date] [Description] [$Amount]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Helper Functions

clean_action(action)

Extract simplified action from verbose Fidelity text:

def clean_action(action):
    if "DIVIDEND RECEIVED" in action:
        return "DIVIDEND"
    if "REINVESTMENT" in action:
        return "REINVESTMENT"
    if "DEBIT CARD PURCHASE" in action:
        return "DEBIT CARD"
    if "MARGIN INTEREST" in action:
        return "MARGIN INTEREST"
    if "DIRECT DEPOSIT" in action:
        return "DIRECT DEPOSIT"
    if "LONG-TERM CAP GAIN" in action:
        return "CAP GAIN"
    if "JOURNALED" in action:
        return "JOURNAL"
    return action[:30]  # Truncate

get_month_name(date_str)

from datetime import datetime
def get_month_name(date_str):
    date = datetime.strptime(date_str, "%m/%d/%Y")
    return date.strftime("%B")  # "January", "February", etc.

format_amount(amount)

def format_amount(amount):
    return f"${abs(float(amount)):,.2f}"

Error Handling

Missing CSV

ERROR: Transaction history CSV not found at notebooks/transactions/
Please download from Fidelity and place in the transactions folder.

Empty CSV

WARNING: CSV contains no transactions to import.

Google Sheets API Error

ERROR: Failed to update Google Sheets. Check:
1. Spreadsheet ID is correct
2. MCP gdrive server is connected
3. You have edit permissions

Validation Checklist

After sync, verify:

• [ ] Transactions tab contains new entries
• [ ] No duplicate transactions added
• [ ] Expense Tracker has new debit card purchases
• [ ] Categories are correctly assigned
• [ ] Uncategorized items are flagged in summary
• [ ] Amounts match CSV values

---

Workflow Type: Data sync Estimated Duration: 30-60 seconds Dependencies: mcp__gdrive__sheets, CSV file access