wenyupapa-sys/tradingagents-codex-skill
tradingagents-codex
VerifiedRun the local TradingAgents Codex fork for multi-agent stock and ETF research. Use when the user asks to analyze tickers with TradingAgents, run codex-analyze, batch stock/ETF research, summarize TradingAgents reports, validate the TradingAgents Codex setup, or turn QQQ/TSLA-style full-chain analysis into a repeatable workflow.
TradingAgents Codex
Overview
Use this skill to run and summarize a local TradingAgents_codex project as an engineering workflow. It wraps setup checks, dry-run command construction, full-chain analysis runs, and report summarization.
Default local project:
~/TradingAgents_codex
Override it with --project or TRADINGAGENTS_PROJECT.
Triggers
Use this skill for requests like:
- "run TradingAgents for QQQ"
- "analyze TSLA with all agents"
- "run codex-analyze"
- "summarize a TradingAgents report"
- "validate the TradingAgents Codex CLI"
- "batch stock/ETF analysis"
- "compare QQQ and TSLA TradingAgents output"
Workflow
- Identify tickers, analysis date, output language, and whether the user wants full-chain or
--analysts-only. - Run setup validation before real analysis unless it was validated in the current session:
SKILL_DIR="${CLAUDE_SKILL_DIR:-${CODEX_HOME:-$HOME/.codex}/skills/tradingagents-codex}"
python3 "$SKILL_DIR/scripts/check_setup.py"
- Dry-run the command when the run is expensive, batched, or date-sensitive:
SKILL_DIR="${CLAUDE_SKILL_DIR:-${CODEX_HOME:-$HOME/.codex}/skills/tradingagents-codex}"
python3 "$SKILL_DIR/scripts/run_analysis.py" \
--ticker QQQ --ticker TSLA \
--date 2026-05-08 \
--output-language English \
--dry-run
- Run full-chain analysis. Tickers run sequentially;
--max-parallelis passed through to TradingAgents for internal role parallelism.
SKILL_DIR="${CLAUDE_SKILL_DIR:-${CODEX_HOME:-$HOME/.codex}/skills/tradingagents-codex}"
python3 "$SKILL_DIR/scripts/run_analysis.py" \
--ticker QQQ --ticker TSLA \
--date 2026-05-08 \
--model gpt-5.5 \
--reasoning-effort high \
--output-language English \
--confirm
- Summarize final portfolio decisions from
reports/portfolio_manager.json:
SKILL_DIR="${CLAUDE_SKILL_DIR:-${CODEX_HOME:-$HOME/.codex}/skills/tradingagents-codex}"
python3 "$SKILL_DIR/scripts/summarize_report.py" \
--ticker QQQ \
--date 2026-05-08
Defaults
- Model:
gpt-5.5 - Reasoning effort:
high - Analysts:
market,news,social,fundamentals - Search: enabled
- Output language:
Chinese - Per-ticker timeout: 3600 seconds
- Latest date fallback: previous likely completed US trading day, skipping weekends
For speed, prefer --output-language English and provide the user a Chinese summary. Chinese output can spend substantial time translating each role report.
Safety
- Treat outputs as research summaries, not investment advice.
- Never place trades, send orders, or connect brokerage execution from this skill.
- Do not invent missing prices, social data, ETF fundamentals, or latest OHLC rows.
- Mention known data gaps, especially ETF fundamentals and missing analysis-date market rows.
- Use
--dry-runbefore large batches. The runner requires--confirmfor more than three tickers. - Always clear
NODE_OPTIONSfor child TradingAgents/Codex CLI commands; this project has a known--use-env-proxyfailure.
Scripts
scripts/check_setup.py: validates project path, virtualenv CLI, required flags, UTF-8 locale, andNODE_OPTIONSworkaround.scripts/run_analysis.py: builds and runstradingagents codex-analyzecommands with dry-run, timeout, and batch confirmation.scripts/summarize_report.py: readsportfolio_manager.jsonand emits Chinese Markdown or JSON summary.scripts/quick_validate.py: validates the skill structure and scripts without running live analysis.
References
Read references/workflow.md for command examples and operational choices. Read references/troubleshooting.md for known failures and mitigations.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |