agentsclimarketplace

Recommendations

Skill mahmoud20138/Tradecraft/plugins/tradecraft/skills/recommendations

Entry-point ranked-setup feed. Runs the full pair-analyze pipeline against a watchlist and returns the top N highest-grade trade setups right now, sorted by confidence × R:R, with one-line rationale and a copy-pasteable trade plan. USE FOR - what should I trade, recommend trades, best setups, top picks, ranked opportunities, where's the edge.From its SKILL.md

Install
npx -y skills add mahmoud20138/Tradecraft --skill recommendations

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 18 stars18 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

5.2 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it

Recommendations — Ranked Setups Entry Point

One of the five entry-point commands for Tradecraft. Returns the top trade setups across a watchlist right now, ranked by quality. Skip the noise — get the picks.

Invocation

/tradecraft:recommendations [SYMBOL,SYMBOL,...] [N]

# examples
/tradecraft:recommendations                              # default watchlist, top 3
/tradecraft:recommendations XAUUSD,US30,EURUSD           # those 3 only
/tradecraft:recommendations BTCUSD,ETHUSD,SOLUSD 5       # crypto top 5
/tradecraft:recommendations all 10                       # default watchlist, top 10
  • [SYMBOL,...] — comma-separated, default watchlist if omitted (same as markets)
  • [N] — number of recommendations to return; default 3

Default Watchlist

XAUUSD, US30, US500, US100, EURUSD, GBPUSD, BTCUSD, ETHUSD

What Claude Should Do

  1. Parse args. Empty symbols → default watchlist.
  2. For each symbol in parallel: run pair-analyze <SYM> H1 standard (or the symbol-specific wrapper if available, e.g. analyze-gold, analyze-us30).
  3. Drop any setup that is bias: flat or fails its news gate.
  4. Score remaining setups: score = setup_grade × confidence × min(risk_reward / 2.0, 1.5) where setup_grade maps A=1.0, B=0.7, C=0.4.
  5. Sort descending; take top N.
  6. Optional: cross-check with ai-signal-aggregator if active for an ML confidence overlay.
  7. Render as the table below.

Output Shape

Top 3 setups — 2026-04-23 14:30 UTC   (default watchlist, mode: standard)

# Rank   Symbol    Side    Entry      SL         TP1        R:R    Grade   Conf   Why
─────────────────────────────────────────────────────────────────────────────────────
1        XAUUSD    long    2341.50    2338.00    2348.50    2.0    A       0.74   H4 trend-up + H1 OB + Asian sweep
2        US30      short   42180      42250      41960      3.1    A       0.68   H4 CHoCH + equal-highs sweep target
3        BTCUSD    long    67200      66400      69000      2.3    B       0.61   D1 breakout retest + funding flat

Skipped: EURUSD (CPI 18h), US500 (range, no edge), US100 (regime + CPI overlap)
Risk note: 2 USD-quote setups; if both fill, total USD exposure = 1.5%.

All numbers above are illustrative — actual values come from the underlying pair-analyze runs.

Output Contract Per Row

ColumnSourceNotes
#rank1 = best
Symbolinputnormalized
Sidepair-analyze.biaslong / short only — flat is filtered out
Entrypair-analyze.entrybroker-precision rounded
SLpair-analyze.stop_lossbroker-precision rounded
TP1pair-analyze.take_profit_1first target
R:Rpair-analyze.risk_rewardTP1-based
Gradepair-analyze.setup_gradeA / B / C
Confpair-analyze.confidence0.0 - 1.0
Whypair-analyze.confluence first 2-3 items joinedone short clause

End the table with a single line listing skipped symbols + reasons, and a portfolio-heat note if multiple setups share an exposure (USD, EUR, risk-on basket).

Skip Conditions

Skip a symbol entirely (don't show as a setup) if any of these fire:

  • pair-analyze.bias == "flat"
  • News gate blocks within next 2 hours
  • Spread > 2× typical for that symbol
  • Regime is high-vol AND mode is not swing
  • Symbol fetch failed entirely (note at bottom, don't include in ranking)

Design Rules

  • Cap at user-requested N or 10, whichever is smaller. Asking for "top 100" is meaningless when most filter out.
  • Never invent a setup — if pair-analyze returns no plan for a symbol, it doesn't appear in the ranking.
  • The Why column is a short clause from confluence, not a paragraph.
  • End with one portfolio-heat note (e.g., "2 USD-quote setups, total USD exposure = X%").
  • If zero setups pass all gates, return: "No A/B-grade setups right now. Next opportunity window: <time of next non-news session>."

Related

  • pair-analyze — underlying per-symbol pipeline this fans out across the watchlist
  • analyze-gold / analyze-us30 — symbol-specific wrappers used when applicable
  • ai-signal-aggregator — optional ML confidence overlay
  • risk-and-portfolio — portfolio-heat math
  • markets — sister entry point for raw quotes (no setup ranking)
  • analyze — sister entry point for deep-dive on one symbol

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 610,677. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.