
from openclaw-retail-trader33
Generate China A-share retail-style trading suggestions and short-term strategies, written in an authentic retail trader voice with actionable entry, exit, and
This skill produces A-share (China) retail-trader style trading suggestions: concise, action-focused plans framed as a retail trader would think. It guides the agent through information intake, setup analysis, persona selection, and structured response building so outputs read like an on-the-spot retail trader recommendation rather than institutional analysis.
Use when the user asks for short-term or swing trading ideas on China A-shares or ETFs and requests a retail-minded perspective (FOMO-aware, emotion-forward). Trigger on explicit phrases like “按 A 股散户思路分析”, “想要 A 股交易建议”, or requests for buy/sell plans, stop loss, and position sizing in a retail voice.
Inferable agents: agents that can read local prompts and references and optionally call market-data skills (eastmoney) for live checks.
A prompt-template skill that instructs an agent to role-play as a China A-share retail trader, generating trading suggestions with an authentic散户 (retail) voice. The SKILL.md is well-written with clear triggers, workflow steps, persona selection, and safety boundaries. No scripts are bundled — it's purely a prompt-engineering skill referencing external prompt templates (prompts/) and a reference playbook (references/) that aren't included in the audit data. Security is solid: it explicitly forbids fabricating prices, promising returns, or recommending leverage. Niche audience — only useful for Chinese A-share traders wanting a retail-perspective simulation.
Prompt-only skill with no executable code. Security score slightly reduced (-12) because the skill instructs the agent to make trading recommendations in a retail-trader voice, which could be misused to give financial advice without disclaimers, though it does include safety boundaries against promising returns. The niche scope (China A-shares, Chinese language only) significantly limits usefulness. Architecture is decent — follows a structured workflow with persona selection and correction handling — but the actual prompt files and references referenced in SKILL.md are not bundled, so a user would need the full repo.