trader-exit

The Exit Planner decides how a trade ends before it starts. The stop is where the setup has failed. The target is where you take what you came for. Both are set BEFORE the entry, written down and handed on, so nobody has to decide anything in the middle of a fast move. A stop is a plan, not a guarantee: a gap or a fast market can fill past it. Day trading loses most people money, and a neat plan does not change that.

${CLAUDE_SKILL_DIR} below is this skill's folder (where this SKILL.md lives).

python3 "${CLAUDE_SKILL_DIR}/exit.py" plan                          # trader/entry.json: midpoint stop, 2R target
python3 "${CLAUDE_SKILL_DIR}/exit.py" plan --stop or-mid --target 1R --write
python3 "${CLAUDE_SKILL_DIR}/exit.py" plan --stop atr:1.5           # needs the session's bars (Yahoo or --offline)
python3 "${CLAUDE_SKILL_DIR}/exit.py" plan --symbol TSLA --entry 364.18 --or-high 362.18 --or-low 354.47
python3 "${CLAUDE_SKILL_DIR}/exit.py" show                          # the current trader/plan.json

Other flags: --flat-by 15:55, --json, --now "2026-10-19 10:15" (New York time), --entry PRICE with a waiting entry.json to plan ahead, --setup level, --offline FILE (a saved Yahoo chart JSON for an ATR stop).

What it reads and writes

Stops and targets

A level setup has no opening range to lean on, so it needs a stop you name.

• Rounding. The stop rounds AWAY from the entry to the cent (down, for a long). Risk = entry minus stop. Target = entry + R x risk, rounded half up to the cent. 2R by default; 1R, 1.5R or price:X also work.
• Refused, with the reason. A stop at or above the entry. A target at or below it. A risk under 0.05% of the price: too tight to survive the spread, normal wiggle would stop you out. An entry.json that has not triggered (unless you give --entry). A short: v2 plans longs only.
• Flat by. 15:55 by default (or flat_by in the config). Out by then, win or lose, no overnight hold. On a 1:00 p.m. early close it moves up to 12:55. An entry at or after the flat-by time is refused.

Steps

1. Run it right after /trader-entry triggers. Before that, --entry PRICE plans ahead at the level you expect.
2. Use the default stop for an opening-range breakout unless the user names another. A level setup needs one: ask.
3. Run exit.py plan --write. Read back the entry, the stop and its rule, the target and its rule, the risk a share, reward:risk and flat by.
4. If it refuses, say the reason and the fix (a wider stop, a farther target). Never loosen anything on your own.
5. Say once that the stop is a plan, not a guarantee.
6. Hand off to /trader-timemachine to replay the plan, then /trader-budget to size it to the max loss per trade. Offer /trader-alerts for the stop, the target and the flat-by time. Claude can keep entry.py watch running in the background or on a /loop until a setup triggers, then come here.

Example: the reel

TSLA, Monday 2026-10-19. The entry is 364.18 at 10:15 AM (the 10:10 bar, bar 8) and the opening range is 354.47 to 362.18. exit.py plan --stop or-mid --target 1R --write:

Exit plan: TSLA long, Mon 2026-10-19, opening-range breakout on 5-minute bars
  Entry         $364.18 at 10:15 AM (bar 8)
  Stop          $358.32  opening-range midpoint: ($362.18 + $354.47) / 2 = $358.325, rounded down
  Target        $370.04  1R (one times the risk)
  Risk a share  $5.86 (1.61% of the entry)
  Reward:risk   1.0
  Flat by       15:55 ET, win or lose: no overnight hold

The default 2R target would be 364.18 + 2 x 5.86 = 375.90. trader-budget then sizes it: a $50 max loss per trade buys floor(50 / 5.86) = 8 shares.

Rules

• The stop and the target are set before the entry, never after.
• Never move a stop further from the entry once the trade is on.
• Plain numbers only: entry, stop, target, risk a share, reward:risk. It never prints a P&L or a return.
• Nothing here promises winning trades or money. A 1R target is not a prediction that the price gets there.
• Long side only. v2 plans longs only.

What it never does

It never places an order, never sizes a position, never widens a stop to make a plan pass, and never calls a stop a guarantee.

Safety

• Education and tooling, not financial advice. A PASS is a rules check, not a recommendation.
• Paper trading by default ("paper": true). Going live is the user's own decision, in their own config.
• Nothing in this pack places an order except trader-broker, and only when all three hold: trader-risk passed that exact order today, the user typed an explicit yes for that one order in the chat, and the user clicked Allow once on Claude's own card because the order tool is set to "Needs approval" (Customize > Connectors > tool permissions). Never set an order tool to "Always allow".
• No options, margin, short selling or leverage unless the config allows it and the user confirms.
• Broker keys live in environment variables or the Claude Desktop config, never in the chat.