long-term-investing-recommendation

name	long-term-investing-recommendation
description	Use when generating long-term stock recommendations from this investing workspace. Runs the long-term recommendation and backtest workflow using the same downloaded daily market data, but treats it as a recommendation-only screener that returns BUY, SELL, or HOLD without managing portfolio state.

Long-Term Investing Recommendation

Use this skill for longer-horizon recommendation runs in this repo.

Working directory

Assume the skill runs from this directory:

/Users/anuragmohan/Documents/POC/investing

If the runtime starts elsewhere, change into that directory first or use absolute paths.

What this skill does

• Runs the long-term backtest and recommendation workflow
• Uses the watchlist from input_config.json
• Does not use or update state/portfolio_state.json
• Produces long-term BUY, SELL, and HOLD recommendations
• Includes long-term backtest context with each run

Commands

Run from the working directory above:

python3 /Users/anuragmohan/Documents/POC/investing/main.py longterm-backtest --format text
python3 /Users/anuragmohan/Documents/POC/investing/main.py longterm-recommend --format text
python3 /Users/anuragmohan/Documents/POC/investing/main.py longterm-run-all --format text

If JSON output is needed:

python3 /Users/anuragmohan/Documents/POC/investing/main.py longterm-run-all --format json --output /Users/anuragmohan/Documents/POC/investing/longterm_output.json

Inputs

• Watchlist, risk preference: /Users/anuragmohan/Documents/POC/investing/input_config.json

The long-term workflow is recommendation-only. It ignores portfolio holdings for recommendation generation.

Strategy style

This workflow looks for slower, durable trend signals such as:

• price above the 200-day moving average
• bullish 50/100/200-day alignment
• positive multi-month momentum
• healthy but not broken trend structure

It is intended for position-style ideas, not short-term swing entries.

Scheduling guidance

Suggested frequency:

• once daily after the close, around 5:00 PM ET

Optional second run:

• once in the morning around 8:00-9:00 AM ET for a refreshed readout

Do not run this every few hours unless there is a specific reason.

Expected output

Summarize:

• strategy mode
• long-term backtest metrics
• BUY, SELL, and HOLD recommendations

For each ticker, include:

• ticker
• action
• confidence
• current price
• short rationale

Guardrails

• Treat this as a screener, not an execution engine
• Do not calculate position sizes unless explicitly requested
• Do not read from or mutate state/portfolio_state.json
• If data fetch fails, report that the recommendation set is incomplete or unreliable
• If backtest mode is blocked, do not elevate BUY ideas