check_watchlist_support
ChatGPTPrimary tool for watchlist-level support and breakdown questions such as 'have any stocks on my watchlist fallen below support?', 'which of my saved stocks broke support?', 'has anything on my watchlist lost support?', 'are any watchlist stocks breaking down?', or 'which watchlist stocks have broken recent structure?'. Checks every saved watchlist ticker against completed EdgeBreak daily chart history. Prefers a genuine published EdgeBreak Launchpad support zone when available; otherwise derives recent price-structure support from existing completed OHLCV. Returns below support, testing support, above support, reclaimed support, or unavailable, and clearly identifies whether support is published or derived. Do not use the plain watchlist action as a substitute for this support/breakdown intent.
edge_journey
ChatGPTPrimary customer-facing EdgeBreak V1 journey for browsing current published Pre-Breakouts, Breakouts and PBI stocks, narrowing an existing result set with supported EdgeBreak master-data filters, rendering branded comparisons for 2–5 named or remaining stocks, opening a Stock Hub, navigating back without losing filters, managing the customer's EdgeBreak watchlist, and answering watchlist-level support/breakdown questions from completed EdgeBreak chart history. Use this tool when the user is exploring EdgeBreak interactively or continuing a scanner workflow in natural language. For comparison intent, prefer this tool over a generic markdown table so the EdgeBreak comparison UI renders. It preserves scanner/date/filter/result/stock state and never reruns or changes scanner formulas. If a requested EdgeBreak V1 action is unsupported or ambiguous, preserve the customer's current state and use closest_supported or continue_journey to surface the nearest valid path rather than dropping context. The customer can always continue normally in ChatGPT. For a standalone visual chart request use show_stock_chart immediately; for detailed company research use get_company_research.
get_company_research
ChatGPTUse for the wider EdgeBreak paid research experience beyond scanner intelligence. Returns one selected research section for a NASDAQ ticker: company overview, financials, institutional research, recent company news, SEC filings, or current market attention. Use get_stock_intelligence for scanner/technical/PBI questions. Use recent_news for news requests and market_attention for attention/discussion requests even though EdgeBreak may research and cache those together internally. Preserve reporting periods and filing dates, do not present lagged institutional disclosures as real-time activity, and do not turn factual research into investment advice.
get_latest_scan
ChatGPTUse for general requests for EdgeBreak's latest or current breakout stocks, strongest current breakout setups, today's scan, leaders today, or what EdgeBreak is seeing today when no more specific intent is present. Returns the latest published NASDAQ breakout scan in EdgeBreak's published order. Present returned multi-stock results in the branded EdgeBreak results card rather than a generic ChatGPT markdown table. Prefer get_pre_breakout_scan for stocks approaching breakout or near resistance, get_pbi_leaders for pressure/PBI requests, search_signals when the user specifies filters, and get_stock_intelligence for a broad assessment of one ticker. Do not invent a new overall ranking beyond published EdgeBreak fields.
get_pbi_leaders
ChatGPTPrimary tool whenever the user's intent is pressure building or PBI: highest PBI, strongest pressure, stocks building pressure, top pressure stocks, strongest build-ups, or PBI leaders. Returns compact leader records for the highest current EdgeBreak Pressure Building Index scores across the full daily scanned universe. Present returned multi-stock results in the branded EdgeBreak results card rather than a generic ChatGPT markdown table. PBI rank is ranking within the PBI universe and is NOT Daily Top 20 or Daily Top 6 rank; when universe_count is available, include that returned count in the customer-facing rank wording rather than hard-coding a universe size or using an ambiguous 'Rank #1'. If the user also asks to see charts, call show_stock_chart for the requested tickers. These are EdgeBreak research records, not investment recommendations.
get_pre_breakout_scan
ChatGPTPrimary tool for stocks approaching a breakout before it happens. Use for natural requests such as 'about to break out', 'close to breaking out', 'getting ready to break', 'near resistance', 'pre-breakout stocks', 'stocks setting up', 'early setups', 'building a base', 'potential breakouts', or 'what hasn't broken out yet'. Returns the latest published EdgeBreak Pre-Breakout Scanner in published order. Present returned multi-stock results in the branded EdgeBreak results card rather than a generic ChatGPT markdown table. For superlative requests such as 'closest to resistance', request enough published results to compare the available distance_to_resistance values before answering; do not invent a new scanner score. Treat published position as list order, not proof of overall quality or liquidity: prefer wording such as 'leading published Pre-Breakout results' rather than 'best' or 'strongest' unless published fields explicitly support that claim. Use search_signals instead when the user is primarily asking for custom filters on the latest breakout scan. If the user asks which pre-breakout names have the strongest pressure, combine published Pre-Breakout results with PBI/intelligence data only as needed. Never expose EdgeBreak methodology.
get_signal_history
ChatGPTUse for a ticker's past EdgeBreak breakout appearances: whether it appeared before, when it showed up, how many times it was picked, or its previous published breakout signals. Returns verified historical EdgeBreak breakout-session records from retained full daily snapshots plus validated exact historical scanner versions. This history may still be incomplete before the published coverage_start date. A zero-match result must NOT be presented as proof that the ticker never appeared outside the verified coverage window. Use the returned session count and coverage dates in customer-facing wording. This is EdgeBreak signal history, not general market-price history; use show_stock_chart for a visual price chart or get_stock_chart_data for raw OHLCV history. If the user asks for past appearances and what the stock looks like now, combine with get_stock_intelligence.
get_stock_chart_data
ChatGPTUse when raw daily OHLCV bars, machine-readable price history, chart-history calculations, or non-visual chart data are actually required. Returns EdgeBreak daily bar history and published chart overlays for one scanned NASDAQ ticker. Do NOT use this as the primary response when the user asks to see, show, chart, graph, visualise, or look at a stock; use show_stock_chart for visual intent. Do not redraw the EdgeBreak chart when show_stock_chart can provide the branded MCP App chart. This tool may also verify current Chart Contract freshness and technical levels when the user explicitly asks for factual levels; summarize only the relevant verified levels rather than dumping raw bars unless requested. Keep factual levels separate from personalised execution guidance.
get_stock_intelligence
ChatGPTDefault tool for broad individual-stock EdgeBreak research. Use for requests such as 'what does EdgeBreak think of SGMT?', 'tell me about SGMT', 'analyse SGMT', 'what are you seeing in SGMT?', 'how does SGMT look?', 'is SGMT setting up?' or 'check SGMT'. The default compact response contains the customer-relevant scanner, structure, momentum/volume, off-exchange, PBI and ranking fields needed for a concise assessment. Set detail_level to full only when the user explicitly asks for full technical detail, all indicators, the full record or a detailed technical table. Use get_stock_signal only when the user specifically wants current breakout-scan membership/status. If the user also asks to see the chart, also call show_stock_chart. Present friendly customer language, keep PBI rank distinct from Daily Top 20/Top 6, and never infer buying or selling direction from off-exchange activity alone. Keep the output factual and setup-focused; do not turn EdgeBreak intelligence into a personalised buy/sell recommendation.
get_stock_signal
ChatGPTUse only when the user is specifically asking whether one ticker is present in the latest published EdgeBreak breakout scan or wants that ticker's current breakout-signal status/data. Examples include 'is SGMT in today's breakout scan?', 'what is SGMT's current breakout signal?', or 'did SGMT trigger today?'. For broader questions such as 'what does EdgeBreak think of SGMT?', 'analyse SGMT', 'how does SGMT look?', or 'is SGMT setting up?', prefer get_stock_intelligence. For previous EdgeBreak appearances, use get_signal_history. Keep the returned stock data factual and do not turn it into a personalised trade recommendation.
master_scanner
ChatGPTUse for custom EdgeBreak scans across the full current published EdgeBreak universe, including combinations that go beyond the smaller in-journey refine menu. Translate natural-language requests into the same Master Scanner filter schema used by the EdgeBreak website. Supports setup type, price, PBI, RSI, resistance distance, structure, volume, moving-average position and recent crosses, OBV, participation, off-exchange activity, Daily Top rankings, MACD, Bollinger Bands, trend and sorting. Also manages the customer's saved Master Scanner scans: list, run, save, change, reset or delete. Saved scans persist normalized rules, not the current result tickers. After a Pattern search/refinement, save this scan should reuse that active search's Pattern + Master Scanner rules when safely representable. Prefer this tool for requests such as 'run a master scan', 'find pre-breakouts with rising OBV and an EMA 8/20 cross in the last 3 days', 'run my saved scan', or 'save this scan'. Results come from the same reusable Master Scanner core as the website; do not recreate or approximate scanner logic in ChatGPT. Treat scanner results as technical screening information and do not convert them into personalised buy/sell instructions.
refine_chart_search
ChatGPTRefine the ACTIVE EdgeBreak chart-search result set without silently restarting from the full universe. Use for follow-ups such as 'now only show PBI over 80', 'only rising OBV', 'within 2% of resistance', 'remove the OBV filter', 'reset filters', 'sort these by PBI', or 'show the best five'. Omit search_id to use the current active EdgeBreak search. Filters are cumulative unless explicitly removed or reset. Refined rows retain the same qualitative match classification, factual explanation and chart-open metadata; do not turn internal similarity scores into customer-facing percentages. For chart requests over the active result set, use chart_page_action first or next to obtain a stable batch of at most 4 symbols without rerunning the search. Refined technical matches remain screening information, not personalised trade recommendations.
search_charts
ChatGPTStart a NEW EdgeBreak chart-pattern or similarity search. Use this when the user begins a fresh request such as 'find ascending triangles', 'find charts like this', 'find more like SGMT', or a new natural-language chart-structure search. For pattern plus technical qualifications, include the existing Master Scanner filters in filters. This creates a search_id for conversational follow-ups. Returned matches use qualitative Very Strong / Strong / Good structural-match classifications with factual reasons and chart-open metadata; do not present internal similarity percentages as customer-facing precision. Do not use this tool for 'now only', 'keep these', 'remove', 'sort these', or other refinements of an existing result set; use refine_chart_search instead. Search results are technical matching evidence, not personalised buy/sell recommendations.
search_signals
ChatGPTUse when the user asks to filter or sort the latest published EdgeBreak breakout scan by explicit conditions such as price, published score, volume ratio, breakout strength, touches, higher lows, rank, grade, price group or setup type. Typical requests include 'stocks under $10 with strong volume', 'setups with at least three touches', or 'filter the scanner'. Present returned multi-stock results in the branded EdgeBreak results card rather than a generic ChatGPT markdown table. If the user says 'strong volume' without giving a threshold, apply the explicit filters they did give and sort by published volume_ratio descending rather than inventing an arbitrary cutoff. Do not use for a plain latest scan, plain PBI leaders, a broad single-stock assessment, or the dedicated Pre-Breakout list. For 'near resistance', 'about to break out', or 'hasn't broken out yet', prefer get_pre_breakout_scan unless the request is specifically to filter the breakout scan. Operates only on already-published EdgeBreak fields and never exposes proprietary methodology.
show_stock_chart
ChatGPTPrimary visual tool whenever the user wants to see a stock or its chart: 'show me SGMT', 'SGMT chart', 'show me the chart', 'graph SGMT', 'visualise SGMT', 'what does the chart look like?', 'show price history', or 'give me the 260-day chart'. Displays up to 260 legitimate daily bars in the premium branded EdgeBreak MCP App chart with price history, volume, moving averages, resistance, latest daily close, any available scanner price shown separately, scanner label, PBI rank, Daily Top 20, Daily Top 6 and Launchpad support when available. Prefer this over get_stock_chart_data for visual requests and do not redraw the chart unnecessarily. Call this tool at most once per ticker in a single user turn and never exceed 4 show_stock_chart calls in one turn; if it already returned chart data for that ticker in the current response, do not call it again. After the call, give a brief acknowledgement so the customer is not left with a silent turn. If found is true and bar_count is greater than 0, treat the branded chart app result as successfully supplied to the host and do not speculate that it was not visible or failed to render. If bar_count is 260, it is accurate to call it the full 260-day chart. If bar_count is between 1 and 259, state the actual bar count and do not call it a full 260-day chart. If found is false or bar_count is 0, clearly say no EdgeBreak daily chart history was returned and do not claim that a chart rendered. If the user asks for chart plus EdgeBreak analysis, also call get_stock_intelligence. Keep chart facts and levels separate from personalised execution guidance.