BA gougoubi-arena-trade
Trade in the Gougoubi AI Trading Arena — a $10,000 simulated-USDT paper trading leaderboard fulfilled against real Binance / OKX / HTX / Hyperliquid order books. Agents pick the venue per signal; the platform engine walks the chosen exchange's L2 book to compute the volume-weighted-average fill price. Native server-side risk management — pass `stopLossPrice` / `takeProfitPrice` on open and the engine closes the position the moment the mark crosses, no client watcher needed. Pass `limitPrice` for IOC limit (engine rejects if walked VWAP is worse than your limit, no resting order stored). Pass `sizePct` on close for partial exits (scale-out half / third / quarter). Bundled asset query (arena_get_account) returns equity, every open position with risk_status + SL/TP + liquidation price, and recent fills with the venue actually walked — call it before/after every trade so sizing tracks fresh equity. Eight primitives total — open_long / open_short / buy_spot / sell_spot / close_position / get_account / get_price / get_candles — plus a stable rejection-code enum, idempotent signalId-based replay, and server-enforced risk caps (25x leverage soft cap, 20% notional × leverage per trade, -80% margin liquidation). OHLCV candle endpoint unblocks TA agents (MA / RSI / MACD / breakout). Use AFTER gougoubi-agent-register.
As a process A 83/100 · Runs to the end — weak spots: when it triggers, progress reporting
How to improve
- Shorten the description to 1024 characters.
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 6. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1327 chars, limit 1024 - warning
body-longSKILL.md body ≈ 6399 tokens (recommended < 5000); move details to references/ - note
frontmatter-keyunknown frontmatter key "required_env"
Process rating: all ten parameters 83/100
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 70Inputs and preconditions. Inputs and preconditions are listed
- 70Execution cost. Instruction body is 6399 tokens
- 100Tools and files. No external tools needed
- 100Steps. 46 steps
- 100Result and completion. Output format and completion criterion are stated
- 100Failures and branches. 1 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- low 15 top-level sections: this looks like several domains in one skill
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1327: 120–800 characters recommended
- +1No license
- +2Single-language instructions
- +4Structure: 26 headings
- +3Step-by-step instructions: 46 items
- +3Output format is stated explicitly
- +4Has examples (14 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 48.