agentleFS
tusharagg1tusharagg1/aifolimizer/.claude/skills/trading-desk/SKILL.md

trading-desk

Skill2 starsChanged 5 months agoMITView on GitHub

End-to-end institutional trade pipeline with a hard Portfolio-Manager approve/reject gate. Use when the user wants a full desk workup before committing real money - "run the full process on X", "desk review for X", "take X through the whole pipeline", "should I actually buy X - do it properly". Chains research → risk gate → pre-trade discipline → trade ticket, and REFUSES to emit a ticket on reject.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No single stage can wave a trade through.

    line 13
  2. Distinctive:

    Before any position-sized entry (not a quick quote check)

    line 20
  3. Very specific:

    Run the adversarial-research skill (bull / bear / consensus) on the ticker.

    line 31
  4. Very specific:

    Run the pre-trade-check skill.

    line 45
  5. Distinctive:

    APPROVE: only if: thesis present, no risk-gate halt, crowding < 70 (or contrarian), pre-trade-check PASS, and reward:risk ≥ 1.5.

    line 49
  6. Distinctive:

    Never fabricate numbers - every figure traces to a tool call (see the data-grounding rule in stock-analysis).

    line 75

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Trading Desk (Analyst → Research → Risk → PM Gate → Execution)

Goal

One orchestrated pass that mirrors an institutional desk: independent analysis, adversarial debate, portfolio-level risk check, behavioral discipline gate, and a final Portfolio-Manager decision that GATES whether a trade ticket is emitted. No single stage can wave a trade through. Output ends in APPROVE + ticket or REJECT + reasons - never a ticket on reject.

When to invoke

  • User wants the complete process on a name before real capital
  • A high-conviction idea needs a disciplined second look
  • Before any position-sized entry (not a quick quote check)

Pipeline (run in order; each stage can veto)

Stage 0 - Mandate (REQUIRED):

  • mcp__aifolimizer__get_profile - capital, account types (never hardcode)
  • mcp__aifolimizer__get_ticker_decision_history (ticker=TICKER, max_decisions=5) + mcp__aifolimizer__get_ticker_reflection (symbol=TICKER, n=3) + mcp__aifolimizer__get_cross_ticker_lessons (max_lessons=3) - load prior decisions BEFORE the desk forms a view. If a prior decision exists and this run flips it, the PM must state explicitly WHY (new data / catalyst / price); never silently contradict a logged decision.
  • Confirm ticker, direction (BUY/ADD/TRIM/SELL), horizon, rough size. If the user can't state a thesis, stop here (REJECT: no thesis).

Stage 1 - Analyst + adversarial research:

  • Run the adversarial-research skill (bull / bear / consensus) on the ticker.
  • Capture: fair-value estimate, key bull driver, key bear risk, base-rate.
  • For US names, fold in mcp__aifolimizer__get_dcf_valuation as a quantitative fair-value anchor.

Stage 2 - Risk gate (HARD veto):

  • mcp__aifolimizer__get_risk_gate_state - if BUYs are halted (drawdown / VIX / loss-streak), a BUY/ADD is auto-REJECTED regardless of conviction.
  • mcp__aifolimizer__get_positioning_signals - crowding ≥70 on a BUY/ADD = veto (defer; negative expected alpha for late entries).
  • mcp__aifolimizer__get_concentration_warnings - if this trade breaches single-name/sector limits, veto or force size-down.

Stage 3 - Pre-trade discipline gate (HARD veto):

  • Run the pre-trade-check skill. If it returns REJECT, the desk REJECTS.

Stage 4 - Portfolio-Manager decision:

  • Synthesize stages 1-3 into one explicit verdict:
    • APPROVE only if: thesis present, no risk-gate halt, crowding < 70 (or contrarian), pre-trade-check PASS, and reward:risk ≥ 1.5.
    • Otherwise REJECT with the specific failing gate(s).
  • State conviction (HIGH/MED/LOW) - drives sizing in the ticket.

Stage 5 - Execution (ONLY if APPROVE):

  • mcp__aifolimizer__get_trade_ticket with the action + conviction → entry zone, stop, exit ladder, sizing.
  • mcp__aifolimizer__log_recommendation to forward-track the call.
  • Optionally mcp__aifolimizer__log_hypothesis to register the thesis with acceptance/invalidation criteria.

Output

DESK VERDICT: APPROVE | REJECT
Ticker | Direction | Conviction
Gate results: research ✓/✗ · risk ✓/✗ · crowding ✓/✗ · pre-trade ✓/✗ · R:R ✓/✗
[If APPROVE] Trade ticket: entry zone, stop, exit ladder, size, account
[If REJECT]  Failing gate(s) + what would have to change to revisit

Rules

  • A veto at ANY hard gate (Stage 2 risk halt, Stage 3 pre-trade REJECT) overrides high conviction. Do not emit a ticket.
  • Never fabricate numbers - every figure traces to a tool call (see the data-grounding rule in stock-analysis).
  • Tax note on the ticket: Non-Reg trims realize 50%-inclusion cap gains.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.