add-etf

Overview

Add a new ETF to the ProphitAI market database. This skill automates classification research (what the ETF tracks, asset class, strategy), matches it to the exact snake_case format used in the DB, and loads all data (prices, holdings, ETF info, dividends).

Workflow

Follow these steps in order. Do NOT skip the confirmation step.

Step 1: Get the ticker

Ask the user which ETF ticker to add (if not already provided).

Step 2: Dupe check

Run the embedded helper script to see if the ticker already exists:

source .venv/bin/activate && python .codex/skills/add-etf/scripts/db_lookup.py check <TICKER>

• If exists: true, warn the user and show the existing classification. Ask if they want to proceed anyway (partial reload) or abort.

Step 3: Web research

Use WebSearch to research the ETF:

• What index/asset class does it track?
• What is the investment strategy? (e.g., sovereign bonds, corporate bonds, equity index, commodities)
• Who is the issuer? (e.g., iShares, Vanguard, SPDR)

Search query: "<TICKER> ETF what does it track asset class strategy"

This determines the appropriate sector/industry/sub-industry classification.

Step 4: DB query

Pull all distinct ETF classification values currently in the database:

source .venv/bin/activate && python .codex/skills/add-etf/scripts/db_lookup.py classifications

This returns the exact snake_case sectors, industries, and sub_industries used for ETFs in the DB.

Step 5: Match

Map the research findings to the exact snake_case values from the DB query.

Matching rules:

• ETF sectors: etf, fixed_income_etfs, commodity_etfs, cryptocurrency_etfs, etc.
• Industries/sub-industries: snake_case (e.g., corporate_bond_etfs, abs_and_mbs, sovereign)
• Always use existing DB values when a match exists. Do NOT invent new classifications.
• If no exact match exists, pick the closest existing value and flag it for the user. If the ETF represents a genuinely new category, propose a new snake_case value following the existing naming convention.

Present the mapping to the user:

Ticker: TLT
Sector:       etf
Industry:     fixed_income_etfs
Sub-Industry: sovereign

Step 6: Confirm

Ask the user to confirm the ticker and classifications before proceeding. Use AskUserQuestion with the mapped values shown clearly.

Step 7: Execute

After confirmation, run the loader:

source .venv/bin/activate && python -c "
from prophitai_data.db.add_etf import load_single_etf
load_single_etf(
    '<TICKER>',
    sector='<matched_sector>',
    industry='<matched_industry>',
    sub_industry='<matched_sub_industry>'
)
"

Key References

Classification Format Examples

Sectors: etf, fixed_income_etfs, commodity_etfs, cryptocurrency_etfs

Industries: fixed_income_etfs, commodity_etfs, equity_etfs

Sub-Industries: corporate_bond_etfs, abs_and_mbs, sovereign, senior_loans, credit