菜单
Redesign TradingBot frontend UX without backend changes. Use when user requests better navigation, dashboard simplification, feature-preserving UI refactor, scan-results-trade workflow emphasis, diagnostics separation, or settings consolidation.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
OpenClaw/operator extension for Schwab trading workflows. Canonical Schwab architecture contract is .cursor/skills/schwab-api/SKILL.md. This file documents OpenClaw-specific tool usage and guardrail behavior.
Scaffold a cockpit dashboard panel end to end: a DTO-backed read endpoint (shared in core/ for local + SaaS parity), a frontend panel module with explicit freshness/error states, and provenance rendering. Use when adding or editing a cockpit lane/panel, an /api/cockpit/* route, core/cockpit_service builders, or any static/ panel that consumes ApiResponse data.
Consistent behavior under stale data, endpoint failures, and fallback use: data-quality labeling, provenance/confidence, circuit-breaker handling, and auto-throttle of risk-increasing orders. Use when working on data_health, fallback logic, circuit breakers, fail-closed risk gates, Provenance, or auto-throttle in execution policies.
Execution intelligence scaffolding: order lifecycle normalization, expected-vs- realized slippage, spread/latency/fill-quality attribution, and policy-driven routing (market vs limit, reprice loop, auto-throttle). Use when working on execution.py order routing, exec quality, slippage, reprice loops, core/execution_policies.py, ExecutionState/blotter, or /api/cockpit/execution/*.
Canonical Schwab endpoint map for the Trading Cockpit: which API surfaces exist, how they are classified (market context / symbol intel / portfolio risk / execution / reliability), their session, latency budgets, fallback rules, and degraded-mode behavior. Use when adding a Schwab endpoint, wiring a provider, defining a cockpit DTO, planning capability gaps, or touching core/endpoint_catalog.py, core/contracts/, or core/providers/.
Closed-loop learning: snapshot every trade decision into a decision packet, run weekly diagnostics (false positives by regime, edge decay by setup, execution drag by condition), and emit advisory tuning proposals for scanner weights and guardrail thresholds. Use when working on decision packets, trade review, outcome attribution, weight/threshold feedback, or core/decision_packet.py / trade_review.py / weight_feedback.py.
| name | front-end-design |
| description | Redesign TradingBot frontend UX without backend changes. Use when user requests better navigation, dashboard simplification, feature-preserving UI refactor, scan-results-trade workflow emphasis, diagnostics separation, or settings consolidation. |
schwab_skill/webapp/static/**.api client usage.Prioritize the primary workflow for Charles Schwab users:
Move setup/connectivity/API management to Settings. Move health/debug/status surfaces to Diagnostics. Keep research tools available, not removed.
Operations (default): Scan controls, scan results, ticker detail panel, trade queue, approvals.Research: Single-ticker research, backtest, SEC compare, full report, strategy chat.Diagnostics: Validation, API error rates, token/quote health, lifecycle/debug outputs.Settings: Schwab connect/onboarding, profile presets, live trading controls, 2FA, billing.