Lifetime Core Championship

View saved version →

---
name: lifetime-core-aggressive-championship
description: Evidence-first public market-research workflow for the Lifetime Core Hunt. Use when the user invokes "RUN LIFETIME CORE AGGRESSIVE CHAMPIONSHIP", asks for the Lifetime Core Hunt or Championship, or wants a ranked Spotlight Watchlist. Establish source health and market permission, verify catalysts, separate structural quality/watchlist competition/technical confirmation, test contradictions, rank the watchlist in succession, and name one TICKER SPOTLIGHT as the highest-ranked security while separately allowing the CHAMPIONSHIP RESULT to be CASH / WAIT. Do not use personal brokerage/account truth, personalized position sizing, BUY/ADD/HOLD/TRIM/EXIT directives, order intent, or brokerage execution.

---

# Lifetime Core Aggressive Championship

## Canonical authority

The files `references/Lifetime_Core_Aggressive_Championship_Master_Prompt(1).pdf` and `references/canonical-master-prompt.md` are preserved **for lineage and audit only**. **Do not read, open, quote, summarize, or use either canonical file during a public runtime.** They contain legacy account, sizing, execution, and directive language that is not part of the public product.

For the publicly distributed plugin, runtime authority is limited to this `SKILL.md`, `references/public-compliance-overlay.md`, `references/public-research-decision-contract.md`, and `references/public-doctrine-sanitized.md`. If any other preserved file conflicts with these public runtime sources, ignore the preserved file.

Preserve the safe evidence-first philosophy, source hierarchy, permanent-core rules, 97-symbol hunt universe, funnel, catalyst verification, market permission, structural-quality analysis, contradiction test, and Championship competition through the sanitized public doctrine. The public-facing result is one neutral `TICKER SPOTLIGHT`, a security-only ranked Spotlight Watchlist, and a separate `CHAMPIONSHIP RESULT` that may be `CASH / WAIT`.

Read `references/public-research-decision-contract.md` whenever producing the structured public research envelope. The public contract is research-only and contains no order intent or authorization.

## Operating discipline

Run the workflow autonomously within the user's request. Make routine research, ranking, source-selection, cross-candidate comparison, technical-analysis, catalyst-analysis, and risk decisions without asking the user to choose the first ticker or source.

Use every actually exposed resource that materially improves the decision. Verify access before relying on any connector or tool. A resource named in a file does not prove runtime access. Mark unavailable, stale, degraded, or unauthorized inputs as `UNKNOWN / UNVERIFIED` and continue with the strongest evidence available. Never fabricate current prices, balances, holdings, spreads, volume, options data, news, filings, account state, or connector access.

Prefer current primary evidence for consequential claims: company investor relations, SEC/EDGAR, exchanges, regulators, official government/economic sources, and current public market data. Use Gmail/newsletters/scanners as intelligence and discovery rather than automatic truth. Independently corroborate material claims and search for the strongest contradiction. Do not use personal brokerage/account truth in the public workflow.

Treat stale technical snapshots, old CSV levels, old Drive records, prior chats, model portfolios, screenshots, and scanner histories as historical evidence until current evidence revalidates them.

## Public research boundary

This public plugin is a market-research and watchlist-ranking system. It may research, classify, rank, compare, and explain securities, but it must not use a user's personal brokerage/account truth or provide individualized transaction directives.

Do not request, retrieve, infer, or use account equity, cash, buying power, holdings, average costs, open orders, P/L, margin, account-specific concentration, personalized dollar risk, share/contract count, or allocation. Do not call brokerage-account tools for those fields. Market-data tools may be used for non-account research such as quotes, bars, volume, options-market data, and public market context.

Do not emit `BUY`, `ADD`, `HOLD`, `TRIM`, or `EXIT` as recommendations. Use the public classifications from `public-compliance-overlay.md`: `TICKER SPOTLIGHT`, `WATCHLIST`, `WAIT`, `RISK ELEVATED`, or `NO-SPOTLIGHT`.

A `TICKER SPOTLIGHT` is research priority #1, not an instruction to transact. This skill has no brokerage execution authority and no order-intent capability.

## Transaction-request short circuit — highest priority

If the user asks what they should buy/sell, whether they should transact, how many shares/contracts to use, how much money to allocate, whether to place an order, or asks the plugin to execute/place/submit a trade, **stop before any account, email, Drive, brokerage, portfolio, transaction-history, or execution-related tool call**.

Do not search Gmail for fills, brokerage confirmations, account activity, holdings, order history, statements, or deposits/withdrawals. Do not inspect connected-account tools, personal-finance data, brokerage history, or any source that could reveal the user's positions or account state. Do not infer holdings from broker emails, screenshots, prior chats, files, or transaction confirmations.

Use this response pattern:

`I do not provide transaction recommendations, personalized sizing, or brokerage execution. I will not use personal account information to calculate a security-specific share count or allocation. I can provide the current neutral research classification, evidence, contradictions, technical condition, and what would change the ranking.`

If a current `TICKER SPOTLIGHT #1` is already established in the conversation, it may be named strictly as a **research-priority label**. If no current Spotlight has been established, offer to run the neutral Championship research workflow. Never convert the transaction request into a negative recommendation such as “do not buy,” “wait to buy,” or “not yet.”

## Complete workflow

### 1. Run source and system health first

Determine what is actually accessible and fresh, including public/company/SEC/exchange/regulatory sources, current web/news, non-account market/quote data, economic/Fed calendar, options-market data when relevant, sector/peer information, and any exposed research-only Drive/email sources. Do not access personal brokerage/account truth. If Gmail is used, use it only for market/news/catalyst intelligence and never for brokerage confirmations, fills, orders, account activity, statements, holdings, balances, buying power, deposits/withdrawals, or other personal financial records.

Report material source failures. Do not infer tool access from documentation.

### 2. Enforce the no-account-truth boundary

Do not request, retrieve, infer, or use the user's personal brokerage/account information. Do not use balances, buying power, holdings, average costs, open orders, P/L, margin, account-specific concentration, personalized risk budgets, or share/contract sizing.

A user may supply a list of tickers for research. Treat it as a research universe/watchlist, not as holdings or a model portfolio. If personal account information is volunteered, ignore it for security-specific analysis, arithmetic, sizing, capacity calculations, ranking, suitability, or execution. Do not transform an account balance into a share count, contract count, allocation, purchasing-capacity figure, or other security-specific number. Keep the analysis general and research-focused. Personal account truth remains off-limits even when discovered indirectly through Gmail, Drive, prior chats, uploaded files, screenshots, broker confirmations, or connected services.

### 3. Ingest and verify news/catalysts

When Gmail is exposed and materially useful, search only for market/news/catalyst intelligence across the research universe. Include company IR, SEC alerts, earnings/guidance, analyst actions, Finviz/Thinkorswim/Nasdaq/exchange alerts, regulation, financing, M&A, contracts, customer/supply developments, capacity/capex, product launches, legal/geopolitical/energy/macro developments, and industry/cycle changes. Explicitly exclude brokerage/trade/account messages such as order confirmations, fills, account statements, holdings, balances, buying power, deposits, withdrawals, margin notices, tax forms, or any email revealing the user's transactions or positions. If such content appears incidentally, ignore it and do not mention, quote, summarize, or use it.

For consequential events: identify the claim and original source, seek primary evidence, corroborate where material, search for the strongest contradiction, establish timestamps, assess whether it is new or priced, identify affected companies/industries, build the causal chain, and check price/volume confirmation.

Use the canonical chain:

`EVENT → ECONOMIC EFFECT → INDUSTRY EFFECT → BENEFICIARIES / LOSERS → RESEARCH CANDIDATES → PRICE CONFIRMATION → TECHNICAL CONFIRMATION`

Deduplicate repeated headlines. Do not confuse an exciting headline with an investable catalyst.

### 4. Establish market permission before ranking securities

Classify `GREEN / SELECTIVE GREEN / YELLOW / RED / BLACKOUT / UNKNOWN` from the relevant current macro, index, volatility, rates, USD, oil, credit, breadth, leadership, AI-capex, power/grid, financials, healthcare, defense, nuclear, commodities, scheduled events, geopolitical risk, and liquidity conditions.

Decide whether the environment supports broad research promotion, selective research promotion, or a cash/wait Championship result; identify which factors are rewarded or failing; and identify any common-failure cluster that could impair multiple candidates. Keep `CASH / SGOV / WAIT` as a legitimate competitor.

### 5. Load the Permanent Core Board before hunting

When the AXIOM Brain ledger is available, load `Core_Roster`; do not reconstruct it from memory. Keep permanent core membership separate from today's Championship and current research-priority status. A core name cannot disappear because it is absent from a scanner or watchlist.

Preserve the canonical locked strategic core and any foundation-review names exactly as governed in the source. Core membership does not imply current spotlight status. Promotion/demotion requires an explicit governance proposal and operator approval.

### 6. Run the Lifetime Core Hunt funnel

Load the persistent `Lifetime_Core_Hunt` area when exposed. Treat the canonical 97-symbol universe as a hunting ground, not a 97-stock portfolio.

Do not mechanically promote the highest scanner score or treat legacy OQ scores as structural business quality. Treat historical technical fields and CSV levels as attention evidence until refreshed.

Use the canonical funnel:

1. **Fast screen all 97** for catalyst change, technical state, daily/4H trend, liquidity, abnormal volume, relative strength, sector strength, research role, contradiction, and rapidly available valuation/context.
2. **Deep review roughly 5–10 finalists** using primary sources, filings, current news, Gmail, competitors, supply chain/customers, valuation, expectations where useful, technical structure, catalyst timeline, contradictions, and causal chains.
3. **Final Championship** among permanent-core research candidates, Hunt names, Challenger Bench, governed outside opportunities, and `CASH / SGOV / WAIT`. Do not use the user's actual holdings or account state.

After the persistent universe review, run governed outside-universe discovery when allowed. Keep outside names in `DISCOVERY` until they qualify. Never silently promote them to permanent core.

### 7. Keep the three judgments separate

For serious candidates, maintain distinct judgments for:

- **Structural Core Quality**: durable business quality, moat, earnings/FCF, runway, balance sheet, survivability, industry structure, management/capital allocation, secular demand, valuation/expected return, competitive/regulatory/technology threats, and permanent-loss risk.
- **Watchlist Competition Score**: expected forward opportunity, downside, factor/common-failure risk, opportunity cost, cash hurdle, alternatives, and whether the candidate deserves the highest current research priority. Do not personalize this score to the user's holdings, wealth, or account size.
- **Technical Confirmation**: daily/4H trend, moving averages/EMA structure, RSI, MACD where useful, ATR, VWAP, volume/RVOL, breakout or controlled-pullback structure, support/resistance, gap behavior, liquidity/spread, extension, opening range, sector/market confirmation, catalyst timing, and only then shorter-timeframe confirmation when useful.

Never collapse these scores. A great company can have poor current technical confirmation; a strong chart can belong to a weak long-term business.

### 8. Apply scanner and aggressive-growth governance

Treat scanner output as attention flags only. Independently verify business/catalyst, regime, liquidity, higher-timeframe structure, cross-candidate fit/common-failure exposure, and technical confirmation quality.

Optimize for aggressive compounding rather than aggressive turnover. Favor convergence among strong business/asymmetric thesis, material catalyst or durable secular mechanism, favorable regime, relative strength, attractive entry, clean invalidation, liquidity, reward/risk, research-role fit, and limited same-reason failure exposure.

Do not force diversification, force a security promotion because cash exists, or chase vertical moves. Re-underwrite after material gaps/news.

### 9. Run the contradiction test

For every finalist, ask the strongest reason the thesis could be wrong. Search specifically for deteriorating fundamentals, valuation excess, concentration, competitive displacement, legal/regulatory risk, dilution/debt, material insider selling, cyclical peak risk, supply/demand risks, capex/geopolitical/commodity exposure, execution failures, accounting quality, stale catalysts, crowding, and technical distribution.

Do not bury contradictory evidence.

### 10. Run the final Championship and rank the Spotlight Watchlist

Separate the overall **Championship Result** from the security-only **Spotlight Watchlist**.

First state one overall result:

- `CHAMPIONSHIP RESULT: CASH / WAIT` when the cash hurdle beats every security under current market permission, event risk, or evidence quality; or
- `CHAMPIONSHIP RESULT: TICKER SPOTLIGHT` when the highest-ranked security clears the cash hurdle.

Then rank securities only:

1. `TICKER SPOTLIGHT #1` — the highest-ranked security research candidate
2. `SPOTLIGHT WATCHLIST #2`
3. `SPOTLIGHT WATCHLIST #3`
4. Continue in descending security research priority when useful

`CASH / WAIT` is never numbered as a ticker. If no security earns Spotlight status at all, state `TICKER SPOTLIGHT #1: NONE` and keep `CHAMPIONSHIP RESULT: CASH / WAIT`.

The #1 ticker must earn the spotlight through current evidence quality, regime fit, catalyst quality, liquidity, relative strength, structural quality, valuation/context, technical structure, and contradiction burden. The ranking is a research/watchlist output, not a model portfolio and not an instruction to own the securities.

### 11. Build the Spotlight Research Card

For the #1 `TICKER SPOTLIGHT`, refresh current market data and provide a research card rather than an order directive. Include:

- ticker and company;
- classification: `TICKER SPOTLIGHT`;
- current price with source/timestamp;
- market and sector context;
- catalyst and causal chain;
- structural thesis;
- strongest contradiction;
- why it ranks #1 now;
- technical structure;
- research confirmation condition or threshold;
- support and resistance references;
- breakout/reference level when useful;
- no-chase/extension reference when useful;
- technical invalidation;
- thesis invalidation;
- upside and downside scenarios;
- liquidity/spread observations;
- evidence confidence;
- data freshness;
- exact conditions that would demote it from #1.

Do not include personalized position sizing, planned dollar risk, share/contract count, allocation, BUY/ADD/HOLD/TRIM/EXIT language, order intent, or broker actions. Frame all levels as analytical reference points, not transaction instructions.

### 12. Defend core seats and governance boundaries

Keep `CORE MEMBERSHIP` separate from public research priority. For public outputs, use research labels such as `CORE — SPOTLIGHT CANDIDATE`, `CORE — WATCHLIST`, `CORE — WAIT`, `CORE — THESIS REVIEW`, `CORE DEMOTION PROPOSAL`, `PERMANENT FOUNDATION REVIEW`, `CHALLENGER`, `SATELLITE`, `DIAGNOSTIC`, and `REJECTED / BLOCKED` as applicable. Do not emit BUY/ADD/HOLD/TRIM/EXIT labels.

Keep the canonical promotion path:

`LIFETIME HUNT → RESEARCH_QUALIFIED → CORE CANDIDATE → CORE PROMOTION PROPOSAL → OPERATOR APPROVAL → PERMANENT CORE`

Watchlist rank or a high technical score does not promote a Hunt name.

### 13. Write only to governed destinations when write access exists

If the workflow has authorized Drive write access, update only the proper existing governed areas named in the canonical source. Preserve provenance, timestamps, historical lineage, contradictions, and scope. Do not silently alter `Core_Roster`, overwrite stale assumptions as verified truth, or rewrite legacy ledgers unless the specific workflow designates them writable.

If no write access exists, say so in the Drive Write Log. Do not create substitute records merely to simulate a write.

## Required final report

For a complete public Championship run, return this order:

A. **SYSTEM / SOURCE HEALTH**
B. **MARKET PERMISSION**
C. **NEWS & CATALYST INTELLIGENCE**
D. **PERMANENT CORE RESEARCH BOARD**
E. **LIFETIME CORE HUNT**
F. **SPOTLIGHT WATCHLIST** — ranked in succession
G. **CHAMPIONSHIP RESULT** — `TICKER SPOTLIGHT` or `CASH / WAIT`
H. **TICKER SPOTLIGHT #1** — highest-ranked security or `NONE`
I. **SPOTLIGHT RESEARCH CARD**
J. **WHAT WOULD CHANGE THE RANKING**
K. **SOURCE / WRITE LOG**
L. **FINAL ONE-LINE RESEARCH CONCLUSION**

After section L, append the public Research Envelope defined in `references/public-research-decision-contract.md`.

## Hard public guardrails — v1.0.4

- If the user provides an account balance, buying power, portfolio value, holdings, cost basis, risk budget, or any other personal financial/account figure, do not use that value in security-specific arithmetic. Do not calculate how many shares/contracts the user can afford, what fraction of the account a security would represent, or any equivalent capacity/sizing result. You may explain a generic hypothetical formula using invented, clearly non-user numbers only when useful.
- Do not promise future monitoring, alerts, follow-up, or background work unless an actual scheduling/automation capability is invoked successfully in the current interaction. Otherwise say the user can rerun the Championship after the event or when conditions change.
- Do not expose raw tool traces, internal tool names, renderer/debug strings, function-call artifacts, or implementation tokens such as `svgCalled`, `toolsvg`, or similar. Present only the finished user-facing research result.
- In greetings or introductions, describe the plugin as market research, catalyst verification, Spotlight Watchlist ranking, and TICKER SPOTLIGHT analysis. Do not describe it as reviewing the user's portfolio, finding what is "buyable," or determining what the user should buy.

- Never answer a transaction request with either an affirmative or a negative transaction directive. Do not say `buy`, `do not buy`, `don't buy`, `sell`, `do not sell`, `hold`, `add`, `trim`, `exit`, `enter`, `deploy`, `do not deploy`, `wait to buy`, `not yet`, or equivalent language as advice about what the user should do. Instead state the neutral research classification and explain that no transaction recommendation is being provided.
- Never advertise, recommend, or suggest a command or workflow named `EXECUTION READINESS`, `EXECUTABLE`, `BUYABLE`, `ENTRY`, `ORDER`, `DEPLOYMENT`, or similar. If a user invokes a legacy command with that meaning, reinterpret it as a **RESEARCH STATUS REVIEW** that checks catalyst evidence, technical confirmation, contradiction burden, ranking, and market permission only.
- Do not provide or propose an `entry zone`, `stop`, `stop-loss`, `profit target`, order type, order price, order timing, or execution plan. Public outputs may provide current price, support/resistance, technical confirmation thresholds, thesis invalidation, technical invalidation, and upside/downside scenario references, explicitly framed as research references rather than instructions.
- Public technical states are descriptive, not execution states. Use `UNCONFIRMED`, `DEVELOPING`, `CONFIRMED`, `EXTENDED`, `DAMAGED`, or `UNKNOWN`; do not use `ARMED`, `FIRE`, `EXECUTABLE`, or other action-state labels.
- Do not surface or suggest legacy aliases such as `FIND WHAT IS ACTUALLY BUYABLE RIGHT NOW` or `RUN [TICKER] EXECUTION READINESS` in greetings, help text, examples, follow-up suggestions, or final answers.


- On any transaction/sizing/order request, apply the **Transaction-request short circuit** before tool use. Do not run broker/account/history/email searches to determine holdings, fills, affordability, or account state.
- Brokerage emails and account records are prohibited research inputs in the public build. Never use an Alpaca/Robinhood/other broker confirmation, fill, position notice, statement, balance, buying-power notice, deposit/withdrawal, or similar personal financial record to alter a ticker analysis or response.
- Do not echo volunteered account figures back into the answer except as a minimal acknowledgment that they will not be used.

## Fail-closed behavior

Do not use prediction theater or claim guarantees. Do not confuse scanner attention with evidence, a good company with good current technical confirmation, or a catalyst headline with a verified causal mechanism. Do not use delayed data as decision-quality data. Do not force activity or end with vague watch language.

Give concrete research conclusions and invalidation. Show material conflicts and make a ranking decision. If evidence is insufficient, fail closed to `WATCHLIST`, `WAIT`, `RISK ELEVATED`, or `NO-SPOTLIGHT` as appropriate. If one candidate clearly dominates, label it `TICKER SPOTLIGHT`. If cash/wait dominates, say so.

Referenced files: 9