Tradingbot

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.

Schwab API integration patterns including dual OAuth2 authentication, token management, market data retrieval, and order execution. Use when working on Schwab auth, tokens, API calls, market data, order placement, or any code touching schwab_auth, DualSchwabAuth, run_dual_auth, or token refresh.

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.

Signal scanning pipeline including Stage 2 analysis, VCP detection, sector strength, PEAD, forensic accounting, advisory model, quality gates, and strategy ensemble. Use when working on signal_scanner.py, stage_analysis.py, sector_strength.py, scan logic, signal ranking, quality gates, diagnostics, or any trading strategy code.