Financial Intelligence Skill

Real-time momentum scoring and market intelligence for 6,500+ stocks and crypto assets. Powered by Banana Farmer — an AI momentum scanner that combines technical analysis, price momentum, and social sentiment into a single 0-100 Ripeness Score.

Backed by 730 days of tracked data across 12,450+ signals with a verified 80% five-day win rate.

Quick Start

Option A — Self-provision a free key instantly (no account needed):

curl -s -X POST "https://bananafarmer.app/api/bot/v1/keys/trial" \
  -H "Content-Type: application/json" \
  -d '{"name": "My Agent", "email": "you@example.com"}'

Save the key from the response. One key per email, instant, no credit card.

Option B — Sign up for a full account: bananafarmer.app/developers

Then:

1. 1. Set your key: export BF_API_KEY=bf_bot_your_key_here (or add to OpenClaw config)
2. Try it: python3 scripts/bf-lookup.py AAPL — you get score, badge, RSI, coil, price action, bull/bear case, and what to watch for

That is it. You are now scanning 6,500+ assets for momentum signals.

---

Prompt Examples

Single Ticker Analysis

Look up any stock or crypto symbol for a full momentum profile: score, badge, RSI, coil pattern, EMA alignment, price action, volatility, scoring breakdown, AI summary, and bull/bear cases.

Example prompts:

• - "What's the momentum on AAPL?"
• "Look up TSLA for me"
• "How's Bitcoin looking?"
• "Check NVDA's score and technicals"
• "Is CRWV ripe?"
• "What's the coil score on AMD?"
• "Pull up the full analysis on SMCI"
• "Is PLTR overbought right now?"

How to run:
CODEBLOCK1

What you get back: Score (0-100), badge (ripe/ripening/overripe/too-late/neutral), current price, 1d and 5d change, RSI with overbought/oversold labels, coil score with breakout flag, EMA 20/50 alignment, 52-week high proximity, volatility metrics, scoring breakdown (technical/momentum/social), key drivers, AI summary bullets, bull case, bear case, and what to watch for.

---

Top Signals / Batch Analysis

See the highest-scoring momentum signals right now — the tickers showing the strongest alignment across technical, price action, and social indicators.

Example prompts:

• - "What are the top signals right now?"
• "Show me the hottest momentum setups"
• "Any ripe signals today?"
• "Top 5 stocks by momentum score"
• "What's ripening that I should watch?"
• "Show me the top 20 signals"
• "Any ripe crypto signals?"

How to run:
CODEBLOCK2

What you get back: Ranked table with symbol, score, badge, 1-day change, 5-day change, and key drivers for each signal. Results are deduplicated by symbol automatically.

---

Portfolio Tracking

Track multiple holdings across accounts. Get a morning-brief-style intelligence report with alerts for RSI overbought/oversold, big daily moves, ripe/overripe signals, risk flags, and P&L calculations.

Example prompts:

• - "Check my portfolio"
• "How are my holdings doing?"
• "Run a portfolio brief"
• "Any alerts on my positions?"
• "How's my aggressive account looking?"
• "What's the P&L on my tech holdings?"
• "Any of my holdings overbought?"
• "Which of my stocks are ripe right now?"

How to run:
CODEBLOCK3

Portfolio file format (portfolios.json):
CODEBLOCK4

What you get back: Market status, data freshness, per-account sections with alerts (overbought, oversold, big moves, ripe/overripe signals, too-late warnings, risk-profile mismatches), and detailed holding lines showing price, score, badge, changes, RSI, shares, cost basis, and unrealized P&L with percentages.

Alert types generated:

• - SIGNAL: holding is ripe or overripe
• CAUTION: holding is too-late (momentum exhausted)
• OVERBOUGHT: RSI above 70 (or 80 for strong warning)
• OVERSOLD: RSI below 30 (potential bounce zone)
• BIG MOVE: more than 5% daily change
• WEEKLY: more than 10% five-day change
• NOTE: high momentum in conservative account

---

Market Overview

Get a bird's-eye view of the market: how many signals are firing by badge, what is trending, new ripe signals, and a narrative summary.

Example prompts:

• - "What's the market doing today?"
• "Give me a market pulse"
• "How's the overall momentum landscape?"
• "How many ripe signals are there right now?"
• "What's trending in the market?"
• "Any new ripe signals today?"

How to run:
CODEBLOCK5

What you get back: Narrative summary, signal counts broken down by badge (ripe, ripening, overripe, too-late, neutral), trending symbols, and newly ripe signals that just crossed the threshold.

---

Performance Tracking / Proof Data

See which signals actually played out: winners and losers with real entry prices, current prices, percentage moves, and milestone returns over multiple time horizons.

Example prompts:

• - "Which signals worked this week?"
• "Show me recent winners"
• "What's the track record look like?"
• "Any big movers from recent signals?"
• "Show me winners from the last 30 days"
• "What percentage of signals won this week?"
• "What were the biggest losers recently?"

How to run:
CODEBLOCK6

What you get back: Winners and losers sections, each showing symbol, percentage change, entry price, current price, and milestone returns (1d, 3d, 5d, 10d). Summary line with calculated win rate for the period.

---

Risk Assessment

Evaluate whether a stock is extended, overbought, or showing risk flags. Combine RSI, badge, coil, and volatility data into a risk picture.

Example prompts:

• - "Is TSLA overbought?"
• "What's the risk on NVDA right now?"
• "Is AMD overripe?"
• "Should I be worried about my SMCI position?"
• "What's the max drawdown on CRWV?"
• "Is this too late to buy PLTR?"
• "Any of the top signals looking overextended?"

How to run:
CODEBLOCK7

What to look for in the output:

• - RSI above 70: overbought warning, watch for pullback
• RSI above 80: strongly overbought
• Badge "overripe": already extended, pullback likely
• Badge "too-late": chasing at this level carries elevated risk
• Max drawdown percentage: historical worst case from entry
• Average daily range: how volatile it trades
• Bear case: the AI-generated downside scenario

---

Comparison Queries

Compare multiple tickers side by side for momentum scores, technicals, and risk profiles.

Example prompts:

• - "Compare AAPL vs MSFT momentum"
• "Which has better momentum: NVDA or AMD?"
• "Look up TSLA, RIVN, and LCID"
• "Compare the big tech names — AAPL, GOOGL, MSFT, META"
• "Which mega cap has the highest coil score?"

How to run:
CODEBLOCK8

What you get back: A formatted comparison table showing score, badge, price, RSI, coil score, EMA alignment, 52-week proximity, scoring breakdown, and volatility side by side. Includes a verdict (strongest/weakest momentum) and risk flags (overbought, coiled for breakout).

---

Watchlist Management

Use the top signals and portfolio tools together to build and track watchlists. Filter by badge to focus on ripening setups that are worth monitoring.

Example prompts:

• - "Add NVDA to my watchlist" (add to your portfolios.json)
• "What's ripening that I should watch?"
• "Build me a watchlist of ripening signals"
• "Update my watchlist with today's top ripening stocks"
• "Track these for me: AAPL, NVDA, AMD, TSLA"

How to run:
CODEBLOCK9

Tip: Use bf-watchlist.py picks for the system's daily curated picks, or create a "watchlist" account in your portfolios.json with shares: 0 and cost_basis: 0 for each symbol. The portfolio brief will show scores, badges, RSI, and alerts without P&L calculations.

CODEBLOCK10

---

Sector and Theme Analysis

Analyze momentum across entire sectors, or drill into specific industry groups.

Example prompts:

• - "Which sectors have the most momentum?"
• "What's the hottest sector right now?"
• "How are the semiconductor stocks doing?"
• "Check the EV sector — TSLA, RIVN, LCID, NIO"
• "Run the FAANG names for me"
• "What's happening in biotech?"
• "Check the momentum on airline stocks"

How to run:
CODEBLOCK11

What you get back: The sectors script groups all top signals by sector (Technology, Healthcare, Financials, Energy, Consumer, Industrials, Real Estate, etc.), shows signal count, average score, heat rating (HOT/WARM/COOL/COLD), ripe signal count, and sector leaders. Use bf-compare.py for side-by-side comparison within a sector group.

---

Historical Context and Win Rates

Query the system's track record and statistical performance data.

Example prompts:

• - "What's the 5-day win rate for ripe signals?"
• "How does 1-day performance compare to 5-day?"
• "What's the average return on signals above 90?"
• "How many signals have been tracked total?"
• "What's the historical data span?"
• "Does patience actually improve win rate?"

How to run:
CODEBLOCK12

Track record reference (from 12,450 signals over 730 days):

Holding Period	Win Rate	Avg Return	Avg Win	Avg Loss
1 day	76.5%	+1.35%	+2.07%	-0.97%
3 days

78.4% | +2.69% | +3.87% | -1.62% |
| 5 days | 79.9% | +4.51% | +6.24% | -2.37% |
| 10 days | 79.4% | +5.40% | +7.54% | -2.86% |
| 1 month | 80.1% | +8.16% | +11.26% | -4.33% |
| 2 months | 79.1% | +9.90% | +13.96% | -5.51% |

Key insight: Win rate starts at 76.5% on day one and climbs to 80.1% by one month. The edge is patience.

---

Alert-Style Queries

Check for actionable conditions across your holdings or the broader market.

Example prompts:

• - "Alert me if any holding goes ripe"
• "Any of my stocks overbought?"
• "Which holdings have RSI below 30?"
• "Are any top signals showing a coil above 70?"
• "What in my portfolio has the biggest move today?"
• "Any too-late warnings on my positions?"

How to run:
CODEBLOCK13

The portfolio brief automatically generates alerts. Look for the ALERTS section, which flags: SIGNAL (ripe/overripe), CAUTION (too-late), OVERBOUGHT (RSI > 70), OVERSOLD (RSI < 30), BIG MOVE (> 5% daily), WEEKLY (> 10% five-day), and risk-profile mismatches.

---

System Health Check

Verify data freshness and market status before making decisions.

Example prompts:

• - "Is the data fresh?"
• "Is the market open?"
• "Check system health"
• "Any data issues right now?"

How to run:
CODEBLOCK14

What you get back: Market status (open, closed, pre-market, after-hours), data freshness (live, recent, stale), and any safety advisory. Always check health before acting on signals — stale data during market hours means something is wrong.

---

Understanding the Data

Ripeness Score (0-100)

The score is a composite of four pillars weighted by their predictive power:

Pillar	Weight	What It Measures
Technical Analysis	35-55%	Chart patterns, RSI, moving averages, coil/spring patterns
Momentum

25-30% | Price velocity in the 1-3% early sweet spot, volume confirmation |
| Social Sentiment | 20-45% | Reddit and X mentions, early buzz detection (1.2-2.0x normal activity) |
| Crowd Intelligence | 0-10% | Crypto only: futures positioning, funding rates |

Higher score means stronger alignment across all pillars. A score of 80 with Technical at 45% and Social at 35% tells a different story than 80 with Technical at 55% and Social at 20% — check the scoring breakdown.

Badge System

Badge	Score Range	What It Means	Action
Ripe	75-89	High conviction setup, strong momentum with favorable entry	Best risk/reward window
Ripening

60-74 | Momentum building but not fully formed | Watch, not act — add to watchlist | | Overripe | 90-100 | Extended, may be due for consolidation or pullback | Caution, tighten stops | | Too-Late | N/A | Already made significant move, chasing carries elevated risk | Do not chase | | Neutral | Below 60 | No significant momentum signal | No edge, stay patient |

Score thresholds for significance: 95+ is rare and highest conviction, 85-94 is strong, 80-84 is actionable.

RSI (Relative Strength Index)

RSI measures momentum on a 0-100 scale:

• - Below 30: Oversold. Price has been beaten down; potential bounce zone. Does not mean "buy" — it means selling pressure may be exhausting.
• 30-70: Normal range. No extreme reading.
• Above 70: Overbought. Price has been running hard; watch for pullback. Does not mean "sell" — strong trends stay overbought for weeks.
• Above 80: Strongly overbought. Higher probability of mean reversion.

Coil Score (0-100)

The coil score measures price compression — how tightly a stock's price is consolidating. Think of it as a spring being compressed:

• - Below 40: Loose. Price is moving freely, no compression buildup.
• 40-69: Moderate compression. Some consolidation, but not yet significant.
• 70+: Coiled. Price is compressed into a tight range. This often precedes a sharp directional move (breakout or breakdown). This is the single most predictive indicator in the system.

A stock with a high coil score AND a ripe badge is the strongest setup: momentum is aligned, and price compression suggests the next move could be significant.

EMA 20 and EMA 50

Exponential Moving Averages smooth price data over 20 and 50 days:

• - Price above both EMAs: Bullish trend — short and medium term aligned upward
• Price above EMA 20, below EMA 50: Short-term bounce in a longer downtrend — proceed with caution
• Price below both EMAs: Bearish trend — momentum is against you
• EMA 20 crossing above EMA 50: Golden cross — potential trend change

Proximity to 52-Week High

A decimal from 0 to 1 representing how close the current price is to its 52-week high:

• - 0.95+ (95%+): Near highs — strong relative strength, but resistance ahead
• 0.80-0.95: Healthy uptrend territory
• Below 0.70: Significantly off highs — check if recovery or further decline

---

Track Record

The system is not new. It has been tracking signals for over two years:

• - 12,450+ signals analyzed across 730 days
• 6,563 unique stocks tracked
• 80% five-day win rate with +4.51% average return
• 76.5% one-day win rate climbing to 80.1% by one month
• Win rate is consistent across score thresholds: 80+ scores all perform between 79-81%

Win Rate by Score Threshold (5-day horizon)

Score Range	Win Rate	Avg Return	Sample Size
80-85	80.2%	+4.60%	3,096
85-90

79.2% | +4.45% | 3,115 | | 90-95 | 79.4% | +4.42% | 3,124 | | 95+ | 80.7% | +4.56% | 3,115 |

The Patience Edge

The data shows holding longer improves outcomes. Day-one win rate is 76.5%. By day five, it is 79.9%. By one month, 80.1%. Average returns scale from +1.35% (1 day) to +8.16% (1 month). The optimal risk/reward window is the 5-to-10-day holding period.

This is not a day-trading system. It catches momentum at 2% instead of 15%, then lets the move develop over days.

---

Error Handling

BFAPIKEY not set

CODEBLOCK15

Fix: Export your API key: export BF_API_KEY=bf_bot_your_key_here. Or add it to your OpenClaw config or .env file.

No signal data available

CODEBLOCK16

Cause: The symbol is not tracked, was delisted, or is a very low-volume OTC stock. Banana Farmer tracks 6,500+ stocks from NYSE and NASDAQ plus popular crypto. Penny stocks and OTC issues may not have enough data for a signal.

Fix: Verify the ticker symbol is correct. Try the exchange-standard format (no special characters). Crypto tickers use their standard symbols (BTC, ETH, SOL).

API timeout or connection error

CODEBLOCK17

Cause: The Banana Farmer API did not respond within 15 seconds. This can happen during high-traffic market opens or if the service is temporarily down.

Fix: Wait 30 seconds and retry. If repeated, check system health with python3 scripts/bf-market.py health. If health also times out, the API may be experiencing downtime.

Rate limiting

The API rate limits depend on your tier: Free (10/min, 50/day), Pro (60/min, 10K/day), Max (120/min, 50K/day). Under normal usage you will not hit these limits. If you do:

Fix: Space out requests. The portfolio script fetches one symbol at a time, so a portfolio of 20 holdings makes 21 API calls (20 lookups + 1 health check). This is well within limits.

Stale data warning

If bf-market.py health reports data freshness as "stale" during market hours, the data pipeline may be delayed. Signals and scores are based on data that refreshes every 15 minutes. Stale data (> 30 minutes old) during open market hours means scores may not reflect current conditions.

Fix: Note the staleness in your analysis. Prices move, but momentum signals are directional and usually remain valid for the session unless there is a major intraday reversal.

403 Forbidden

CODEBLOCK18

Cause: Missing or malformed User-Agent header. The API requires a User-Agent: BananaFarmerBot/1.0 header.

Fix: The scripts set this automatically. If you are calling the API directly, make sure to include the header.

---

Advanced Usage

JSON Output Mode

For programmatic processing, the portfolio script supports JSON output:

CODEBLOCK19

This returns a JSON object with a brief field (the formatted text) and a signals field (score and badge for each looked-up symbol). Use this for piping into other tools, dashboards, or automated workflows.

Multi-Account Portfolios

The portfolio file supports multiple accounts with different risk profiles. Each account gets its own section in the brief with account-specific alerts. A conservative account holding a high-momentum stock will get a NOTE alert that an aggressive account would not.

Supported risk profiles: conservative, moderate, aggressive. The --account filter accepts partial matches on both the account id and name fields.

CODEBLOCK20

Combining Scripts

Chain scripts together for richer analysis:

CODEBLOCK21

Filtering Top Signals

The top command supports badge and limit filters:

CODEBLOCK22

Movers Time Range

Control the lookback window for performance tracking:

CODEBLOCK23

---

Scripts Reference

Script	Purpose	Key Arguments
INLINECODE23	Deep analysis of specific tickers	INLINECODE24
INLINECODE25

Market overview and signal scanning | health, top [--limit N] [--badge X], pulse |
| bf-portfolio.py | Portfolio intelligence with alerts | FILE.json [--account NAME] [--json] |
| bf-movers.py | Winners/losers proof data | [--days N] [--limit N] |
| bf-compare.py | Side-by-side ticker comparison table | SYMBOL1 SYMBOL2 [SYMBOL3 ...] [--json] |
| bf-watchlist.py | Curated picks, scorecard, horizons | picks, scorecard, horizons [--json] |
| bf-sectors.py | Sector momentum breakdown | [--json] |

All scripts are in the scripts/ directory. All require python3 and BF_API_KEY in the environment. No additional pip packages are needed — everything uses the Python standard library.

---

Pricing

Plan	Price	What You Get
Free	$0	Health, discover, top 3 signals. 10 req/min, 50/day. Enough to verify it works.
Pro

$49/month ($39/mo annual) | Full 50+ leaderboard, all endpoints, proof images, portfolio, movers, watchlist, 30-day score history. 60 req/min, 10K/day. |
| Max | $149/month ($119/mo annual) | Everything in Pro + historical scores with exact prices at signal, calculated returns, full 730+ day backtesting, bulk export, webhooks. 120 req/min, 50K/day. |

Get your key instantly at bananafarmer.app/developers. Free tier works immediately — no credit card needed.

For comparison: Danelfin Pro charges $79/mo for AI scores with historical data but no prices attached. Polygon.io charges $79-500/mo for raw price data with zero intelligence. Alpha Vantage is $50-250/mo for raw data. Banana Farmer Max at $149/mo gives you both — momentum intelligence AND exact prices at every signal — with 730+ days of backtesting proof. Still less than Polygon's mid-tier, with far more intelligence.

---

Security

This skill is designed with transparency and safety in mind:

• - Outbound HTTPS only: All scripts make only outbound HTTPS calls to bananafarmer.app. No other network connections, no inbound listeners, no file exfiltration.
• Zero pip dependencies: Every script uses only the Python standard library (json, urllib, ssl, os, sys). No third-party packages to audit.
• MIT licensed: Full source code is readable and auditable.
• No secrets in code: API key is read from the BF_API_KEY environment variable only. Never hardcoded, never logged.
• Read-only: The skill reads market data. It does not execute trades, manage accounts, or modify any files on your system.
• Infrastructure security: Security practices — TLS 1.3, AES-256, Cloudflare WAF, Stripe PCI DSS Level 1.
• Legal: Terms of Service · Privacy Policy · System Status

---

Disclaimer

This skill provides financial data, momentum scores, and analytical intelligence. It is not financial advice. All data is for informational and research purposes only.

• - This tool does not make buy or sell recommendations
• Past performance does not guarantee future results
• Users should do their own research and consult a licensed financial advisor before making investment decisions
• Win rates and return figures are historical and based on backtested signal data
• Stock data is delayed 15 minutes per exchange rules; crypto data is near real-time

By using this skill, you agree to the Banana Farmer API Terms.

Market data sourced by Tiingo.com. Momentum scoring, analysis, and the Ripeness Score methodology by Banana Farmer.