Get a current price snapshot for a stock, with 52-week range context.

Informational (not a buy/sell signal). Use this to anchor any analysis with live price, daily change, market cap, volume, and where price sits inside its 52-week range (0% = at the low, 100% = at the high).

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown report or JSON with fields: price, previous_close, open, day_low, day_high, year_low, year_high, market_cap, volume, currency, change_pct, pct_of_52w_range.

Examples: - "What's Apple trading at?" -> ticker='AAPL' - "Is NVDA near its 52-week high?" -> ticker='NVDA'

Measure net insider buying vs. selling from SEC Form 4 filings (last 180 days).

Aggregates open-market Purchase (P) and Sale (S) transactions by USD value and returns a directional score from -100 (heavy net selling) to +100 (heavy net buying). Insider buying is among the more predictive public signals. If a company has no qualifying Form 4 activity, returns "no signal" rather than a fabricated score.

Requires the EDGAR_IDENTITY env var (any 'Name email@example.com') per SEC fair-access policy.

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown or JSON with score, confidence, and data: net_usd, buys_usd, sells_usd, n_filings, net_selling, window_days.

Examples: - "Are insiders buying Tesla?" -> ticker='TSLA' - "Insider sentiment for PLTR" -> ticker='PLTR'

Check which renowned value investors hold a stock and their recent activity.

Reads Dataroma's tracking of ~80 superinvestors' 13F portfolios. Returns a directional score blending ownership breadth (how many hold it) with recent net buying/selling. A widely-held, recently-bought name scores positive.

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown or JSON with score, confidence, and data: count_owners, recent_buys, recent_sells, avg_hold_price, owners_sample.

Examples: - "Do any famous investors own Berkshire?" -> ticker='BRK.B' - "Smart-money interest in GOOGL" -> ticker='GOOGL'

Get Wall Street buy/hold/sell consensus for a stock (Finnhub).

Returns a directional score from the distribution of analyst ratings. Requires FINNHUB_API_KEY; returns "no signal" if unset or no coverage.

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown or JSON with score, confidence, and data: strong_buy, buy, hold, sell, strong_sell, analysts_count, period.

Examples: - "What do analysts think of AMD?" -> ticker='AMD' - "Is MSFT a buy according to Wall Street?" -> ticker='MSFT'

Estimate fair value and financial health for a stock (Yahoo Finance).

Computes a fair-value range using forward EPS × sector P/E band, then measures balance-sheet health (debt/equity, current ratio, gross margin). Score: positive = stock trades below fair value, negative = expensive. A cheap-but-fragile balance sheet halves any positive score.

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown or JSON with score, fair_value, upside_pct, tag, health_score, and underlying fundamentals.

Examples: - "Is KO undervalued right now?" -> ticker='KO' - "What's the intrinsic value of AAPL?" -> ticker='AAPL'

Get the 1-month implied move and put/call OI skew for a stock (yfinance).

The implied move (straddle / spot) is primarily a risk-sizing tool — it tells you how much the market expects the stock to move before the nearest ~30-day expiration. The directional score from put/call OI skew is LOW conviction (options flow is noisy); treat this as a weak secondary indicator only.

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown or JSON with implied_move_pct, expiry, pc_oi_ratio, score.

Examples: - "How volatile is NVDA expected to be?" -> ticker='NVDA' - "What does options flow say about AAPL?" -> ticker='AAPL'

Run a full multi-source analysis of a stock and return one scored verdict.

Fans out to every available source in parallel (price context, insider Form 4 activity, superinvestor holdings) and blends the directional signals into a confidence-weighted verdict from -100 (Bearish/AVOID) to +100 (Bullish/BUY), with a per-source breakdown. Sources that have no data are reported as "no signal" and excluded from the score rather than guessed.

This is the primary tool — prefer it for "should I look at X?" questions; use the individual tools when you only need one dimension.

Args: params: ticker (str) and response_format ('markdown'|'json').

Returns: str: Markdown verdict with gauge, action, signal-breakdown table, and each source's one-line takeaway; or JSON with the full verdict + raw signals.

Examples: - "Give me a full read on Microsoft" -> ticker='MSFT' - "Should I be looking at PLTR?" -> ticker='PLTR'