cash-deployment
Skill2 starsChanged 5 months agoMITView on GitHub
Plan how to deploy uninvested cash across existing portfolio holdings. Use when the user asks "where do I put my cash?", "I have $X to invest", "deploy my cash", "what should I buy with my settled funds?", or "add to my best names". Routes capital to existing tickers ranked by setup quality and avoids concentration pile-on. Fetches portfolio + technicals + fundamentals via aifolimizer MCP.
In this repository29 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- adversarial-researchSkill
- auto-rebalanceSkill
- daily-briefingSkill
- dividend-strategySkill
- earnings-analyzerSkill
- earnings-postmortemSkill
- employer-stockSkill
- health-checkSkill
- macro-impactSkill
- momentum-scannerSkill
- optimize-allocationSkill
- pead-trackerSkill
- perf-optimizerSkill
- portfolio-healthSkill
- portfolio-reviewSkill
- position-reviewSkill
- pre-trade-checkSkill
- profile-setupSkill
- risk-assessmentSkill
- sector-rotationSkill
- stock-analysisSkill
- stock-compareSkill
- tax-loss-reviewSkill
- top-trades-todaySkill
- trade-journalSkill
- trading-deskSkill
- weekly-mirrorSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
20 rules. About 0 are common, 7 are distinctive, and 13 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
13
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Call mcp__aifolimizer__get_cross_ticker_lessons with max_lessons=3 - surface recent stop-outs or wins.
- Very specific: line 12
Call mcp__aifolimizer__recall_preferences with query="cash deployment add position" - investor preferences on sizing and account allocation.
- Very specific: line 13
Call mcp__aifolimizer__get_personal_context - province, marginal tax rate, account waterfall, FHSA/FTHB horizon.
- Very specific: line 15
Call mcp__aifolimizer__get_profile - per-account cash balances.
- Very specific: line 17
Call mcp__aifolimizer__get_portfolio with account_id=<that account> - current holdings, weights, cost basis
- Very specific: line 18
Call mcp__aifolimizer__get_concentration_warnings - already-overweight names to EXCLUDE from add candidates
- Very specific: line 19
Call mcp__aifolimizer__get_fundamentals on existing holdings - sort by valuation + analyst target upside
- Very specific: line 20
Call mcp__aifolimizer__get_technicals on existing holdings - sort by Minervini stage 2 + RSI not overbought + trend uptrend
- Very specific: line 21
Call mcp__aifolimizer__get_positioning_signals on existing holdings - crowding score per name.
- Distinctive: line 53
Not flagged for concentration
- Very specific: line 56
No lottery/MAX flag (lottery_flag != true) - abnormal single-day spike = chase risk (Bali-Cakici-Whitelaw)
- Very specific: line 64
Check get_concentration_warnings - exclude overweight names FIRST
- Distinctive: line 77
If user wants to deploy into different account type, flag contribution-room cost (TFSA 2026 limit, RRSP earned-income headroom)
- Distinctive: line 90
Always render eligible-candidates table (Setup Score /8, Technical Score, Entry, Stop columns required)
- Distinctive: line 91
Never recommend adding to concentration-flagged position
- Distinctive: line 92
Never recommend stage 3 or stage 4 ticker even if user holds it - say "do not add" explicitly
- Very specific: line 96
After plan is finalized: call mcp__aifolimizer__log_trade_decision for each ticker in the deployment plan (parallel calls).
- Distinctive: line 107
Don't double-count: if user has a recurring auto-deposit into a core ETF, mention that flow before recommending another lump add to the …
- Distinctive: line 110
If ALL eligible candidates consensus-crowded, recommend partial deploy (cap 50% of cash) + cash hold for pullback.
- Very specific: line 114
If get_ticker_decision_history shows a recent stop-hit on a candidate, require explicit new bullish evidence (earnings beat, technical …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Cash Deployment (Add-to-Winners with Discipline)
How to run
Step 0 - Memory (call in parallel with Step 1):
- Call
mcp__aifolimizer__get_cross_ticker_lessonswithmax_lessons=3- surface recent stop-outs or wins. If any top add candidates were recently stopped out, require new data to justify re-entry. - Call
mcp__aifolimizer__recall_preferenceswithquery="cash deployment add position"- investor preferences on sizing and account allocation. - Call
mcp__aifolimizer__get_personal_context- province, marginal tax rate, account waterfall, FHSA/FTHB horizon. This grounds the §5 tax-account routing (US-dividend→RRSP, growth→TFSA/FHSA). Ifpresent=false, flag the tax reasoning as generic and suggest the user run profile-setup.
- Call
mcp__aifolimizer__get_profile- per-account cash balances. Cash sits in specific accounts (TFSA, RRSP, Non-Reg) - can't move cash across account types without contribution-room consequences - Confirm with user: which account(s) is cash in? If unstated, use account(s) with largest cash balance from
get_profile - Call
mcp__aifolimizer__get_portfoliowithaccount_id=<that account>- current holdings, weights, cost basis - Call
mcp__aifolimizer__get_concentration_warnings- already-overweight names to EXCLUDE from add candidates - Call
mcp__aifolimizer__get_fundamentalson existing holdings - sort by valuation + analyst target upside - Call
mcp__aifolimizer__get_technicalson existing holdings - sort by Minervini stage 2 + RSI not overbought + trend uptrend - Call
mcp__aifolimizer__get_positioning_signalson existing holdings - crowding score per name. EXCLUDE consensus-crowded names from add candidates - For the top 3 candidates after filtering, call
mcp__aifolimizer__get_ticker_decision_history(one call per candidate, parallel) - check for prior stop-outs or failed entries on these specific tickers - Cross-reference: ideal add = (not concentration-flagged) AND (stage 2 uptrend) AND (RSI 40-65) AND (analyst upside >10%) AND (fundamentals not deteriorating) AND (crowding_score < 70) AND (lottery_flag != true) AND (no recent stop-out without new thesis override)
Investor profile
- Canadian retail investor
- Time horizons: short-term + long-term (10yr+)
- Account types and capital: always read from
get_profile- never hardcode - Strategy lens: confirm with user (defensive top-up / aggressive growth / dividend reinvest); default "aggressive growth" if unstated
Output structure
1. Cash summary (top)
- Account: which account, how much cash settled
- Strategy lens assumed
- Number of candidate positions reviewed
2. Disqualified buckets (one-liner each)
- Concentration-flagged (already overweight) - list tickers, current weight, threshold breached
- Technically broken (stage 3/4, downtrend) - list tickers, why excluded
- Overbought (RSI >70) - list tickers, wait-for-pullback prices
- Fundamentally deteriorating (analyst downgrades, margin compression) - list tickers, what changed
- Consensus-crowded (crowding_score ≥ 70) - list tickers, score, "edge already priced - defer add"
3. Eligible add candidates (markdown table)
Columns: Ticker | Current Weight | Setup Score /8 | Technical Score | Crowding | Entry | Stop | Notes. Setup Score rubric (1 point each, 8 criteria, max 8):
- Stage 2 uptrend
- RSI 40-65 (room to run)
- Above SMA50
- Analyst target upside >10%
- Not flagged for concentration
- Crowding score < 70 (not consensus-crowded)
- Volume score ≥ 1.0 (above-average volume confirmation)
- No lottery/MAX flag (
lottery_flag != true) - abnormal single-day spike = chase risk (Bali-Cakici-Whitelaw)
4. Deployment plan (risk-first)
For the top 3 add candidates (by Setup Score; tiebreak higher technical_score), call mcp__aifolimizer__get_trade_ticket with ticker, action="ADD", conviction=<from Setup Score: 7-8=HIGH, 5-6=MED, 3-4=LOW, <3=skip>. This is the single source of truth for levels (same engine as pre-trade-check) - supersedes raw pivot_levels:
entry_zone-{timing: buy_now | wait_pullback, low, high, reference, support_basis}. Ifwait_pullback, do NOT deploy at market - state the pullback band and route that cash to the next rankedbuy_nowcandidate or hold it.stop_loss_price,exit_ladder(T1/T2/T3 scale-out),positionblock (avg cost / return / stop_below_cost since these are held names).
Before sizing any position:
- Check
get_concentration_warnings- exclude overweight names FIRST - Max position loss = (
entry_zone.reference-stop_loss_price) × shares. Ensure no single add risks >2% of total portfolio - Entry =
entry_zone.reference(only deploy whentiming == buy_now) - Stop =
stop_loss_price - Position size = min(risk-based shares, 5% of total portfolio in dollar terms)
- Dollar allocation per ticker (sum to cash balance, round to whole shares)
- Total share count + estimated execution price; render each name's
exit_ladderas the profit-taking plan - Cash remaining after deployment (if any - keep for opportunistic adds, or for
wait_pullbacknames) - Max loss per position stated explicitly
5. Tax-account reasoning
- Confirm deploy stays in same account as cash (no contribution-room hit)
- If user wants to deploy into different account type, flag contribution-room cost (TFSA 2026 limit, RRSP earned-income headroom)
- US-dividend stocks → prefer RRSP cash (no 15% withholding)
- High-growth no-dividend → TFSA cash preferred (gains tax-free)
- Canadian dividend payers → Non-Reg or TFSA (Cdn dividend tax credit in Non-Reg)
6. What's NOT in this plan
- New tickers outside user's existing universe (use stock-analysis for individual research)
- Bond / GIC / HISA cash parking (deploy ≠ park)
- Crypto allocations (only state this if
get_profileshows no crypto account/holdings; if the user DOES hold crypto, drop this line or address crypto deployment explicitly - never assert it as a static fact)
Rules
- Under 700 words
- Always render eligible-candidates table (Setup Score /8, Technical Score, Entry, Stop columns required)
- Never recommend adding to concentration-flagged position
- Never recommend stage 3 or stage 4 ticker even if user holds it - say "do not add" explicitly
- Whole-share counts only (Wealthsimple supports fractional but plan should be real trade ticket)
- Reference user's actual cash balance from
get_profile- never hardcode - Risk-first: compute max loss per position BEFORE sizing. Never size first and check risk second
- After plan is finalized: call
mcp__aifolimizer__log_trade_decisionfor each ticker in the deployment plan (parallel calls). action=BUY, conviction based on Setup Score (7=Strong Buy, 5-6=Buy, 3-4=Neutral), skill_used="cash-deployment"
Gotchas
get_profilecash balance is per-account - DO NOT sum cash across accounts for single-account deploy. TFSA cash can't buy RRSP positions without in-kind transfer that resets contribution roomget_concentration_warningsthreshold defaults: single 10%, sector 35%. Adjust via tool params if user has different risk tolerance - don't recommend adds above active thresholdsget_technicalscached 1h - entry prices stale on high-volatility days. State timestamp- "Add to winners" can morph into momentum-chasing - if all top candidates extended (>20% above SMA50), recommend partial deploy + cash hold for pullback rather than full deploy
- Settled cash vs unsettled: Wealthsimple T+1 settlement on equity sales. If user just sold, cash may not be available - note in plan
- USD cash in CAD account triggers FX conversion at WS spread (~1.5%). If cash is USD and buy is .TO, flag FX cost
- "Aggressive growth" tilt amplifies single-name risk - cap single add at 5% of total portfolio even if cash > 5%
- Don't double-count: if user has a recurring auto-deposit into a core ETF, mention that flow before recommending another lump add to the same fund
- Tax-loss-harvested cash: superficial-loss rule blocks rebuying same security for 30 days - verify with tax-loss-review before recommending same ticker back
- Crowding score is tilt, not veto for existing holdings - never recommend SELLING held name because it became consensus. Only use to defer/reduce ADDS
- If ALL eligible candidates consensus-crowded, recommend partial deploy (cap 50% of cash) + cash hold for pullback. Do not force full deploy into crowded names
pivot_levelsnull for very new/halted symbols - fall back to current price as entry, SMA50 as stopvolume_scorenull for some TSX ETFs - skip volume criterion (score out of 7 for those tickers, note it)technical_scoreis a composite screening signal; always verify against individual sub-criteria before committing capital- If
get_ticker_decision_historyshows a recent stop-hit on a candidate, require explicit new bullish evidence (earnings beat, technical breakout, crowding unwound) before recommending re-entry. State why this time is different in the Notes column. - Cross-ticker lesson: if get_cross_ticker_lessons shows recent stop-outs in same sector as candidate, flag "sector risk - recent stop pattern" in Notes column
log_trade_decisioncalls are not optional - they build the feedback loop that improves future cash deployment plans
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

