Sandbox
@BaggaT236/AI-Trading-Skills

Claude trading skills for repeatable market workflows

This repo packages trading-focused Claude skills, command files, and a TypeScript runtime for running and validating analysis workflows. It ties the skills catalog to a canonical registry and validation scripts so builders can keep their process consistent. Redis is optional for cached state.

121 stars960 forksPythonUpdated 14d ago
Who it's for

Builders who want Claude Code to follow a disciplined trading workflow for analysis, screening, and review.

What it delivers

You can run repeatable trading analysis with validated skills instead of improvising each session.

What it does

Skill catalog

Provides many reusable trading skill packages for regime analysis, screening, sizing, postmortems, and related market tasks.

Workflow CLI

Includes a strict TypeScript CLI for listing workflows, running workflow jobs, and validating the registry.

Metadata validation

Checks skills, indexes, and workflow definitions before you rely on them.

Optional Redis cache

Supports Redis-backed state caching for faster repeated runs, with a no-Redis mode too.

Command and prompt files

Uses Claude-facing repo files such as `CLAUDE.md` and `commands/` to shape how the agent works in this project.

How to get it

  1. 1Run
    git clone https://github.com/tradermonty/claude-trading-skills.git
    cd claude-trading-skills
    npm install
    cp .env.example .env
    npm run validate

README

Claude Trading Skills

Build disciplined, repeatable trading workflows with a modern TypeScript platform.

This repository gives you:

  • a strict TypeScript CLI for workflow execution and metadata validation,
  • a large skill catalog for market analysis, screening, and review loops,
  • optional Redis-backed state caching for faster repeated runs.

Designed for discretionary traders who want better process quality, not "black-box" auto trading.

Educational use only. Not financial advice.


Why This Repo Stands Out

  • TypeScript-first runtime: strongly typed CLI, validators, config, and tooling.
  • Skill-rich ecosystem: reusable modules for regime analysis, idea generation, sizing, and postmortems.
  • Workflow orchestration: YAML-driven playbooks with clear decision gates.
  • Production hygiene: strict lint/typecheck/test pipeline and deterministic builds.
  • Flexible operation: runs locally with or without Redis.

Quick Start (3 Minutes)

git clone https://github.com/tradermonty/claude-trading-skills.git
cd claude-trading-skills
npm install
cp .env.example .env
npm run validate

If validation passes, you are ready.


First Commands To Run

npx trading-skills list
npx trading-skills workflows
npx trading-skills workflow market-regime-daily
npx trading-skills validate
npx trading-skills validate-index --strict-workflows

Project Layout

claude-trading-skills/
├── src/                  # TypeScript runtime, CLI, config, validators
├── tests/                # Vitest suites
├── skills/               # Skill content and execution assets
├── workflows/            # YAML workflow definitions
├── skillsets/            # Installable bundles by use-case
├── skills-index.yaml     # Canonical skill registry
├── docs/                 # Architecture and contributor docs
└── package.json          # Build/test/lint scripts

TypeScript Development

npm run dev
npm run typecheck
npm run lint
npm run test
npm run build
npm run validate

npm run validate is the canonical local gate before pushing.


Environment Configuration

Copy .env.example to .env and adjust:

  • LOG_LEVEL (info, debug, error)
  • REDIS_ENABLED (true or false)
  • REDIS_URL (defaults to local Redis)
  • REDIS_KEY_PREFIX
  • REDIS_CACHE_TTL_SECONDS
  • SKILLS_DIR
  • SKILLS_INDEX_PATH
  • WORKFLOWS_DIR

No Redis? Set REDIS_ENABLED=false and continue.


Recommended Workflows

  • Daily market posture: market-regime-daily
  • Swing scan routine: swing-opportunity-daily
  • Weekly portfolio review: core-portfolio-weekly
  • Trade journaling loop: trade-memory-loop
  • Monthly performance review: monthly-performance-review

Workflow truth source lives in workflows/ and skills-index.yaml.


Troubleshooting

  • Type errors after pulling: run npm install then npm run typecheck.
  • Lint failures: run npm run lint and fix reported files.
  • Workflow validation errors: run npx trading-skills validate-index --strict-workflows.
  • Redis unavailable: set REDIS_ENABLED=false in .env.

Contributing

  1. Create a feature branch.
  2. Make focused changes.
  3. Run npm run validate.
  4. Update relevant docs when behavior changes.
  5. Open a PR with clear test evidence.

For deeper architecture context, see docs/STRUCTURE.md and docs/AUDIT.md.


FAQ

Is this an autonomous trading bot?
No. It structures analysis and decision quality; humans remain in control.

Can I run without Redis?
Yes. Redis is optional.

What is the canonical registry?
skills-index.yaml.


License

MIT. See LICENSE.

Files in the repo

Repository payload30 top-level entries
  • .github
  • agents
  • audits
  • commands
  • docs
  • examples
  • launchd
  • scripts
  • skill-packages
  • skills
  • skillsets
  • src
  • tests
  • workflows
  • .env.example
  • .gitignore
  • .secrets.baseline
  • CLAUDE.md
  • conftest.py
  • eslint.config.js
  • LICENSE
  • package.json
  • PROJECT_VISION.ja.md
  • PROJECT_VISION.md
  • README.ja.md
  • README.md
  • skills-index.yaml
  • tsconfig.json
  • tsup.config.ts
  • vitest.config.ts

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More skills

Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.

117k
1 add

Open-source AI job search: scan job portals, evaluate listings into a structured A-H report with a global 1-5 score, tailor your CV, track applications — runs locally in your AI coding CLI (Claude Code, Codex, OpenCode, Antigravity…)

71k
ayghri/
i-have-adhd

A skill to stop your coding agent from burying the answer. ADHD-friendly output.

38k
mvanhorn/
last30days-skill

AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary

62k
Nanako0129/
sepia

De-AI writing skill for any Agent Skills-compatible agent (77+ via the Skills CLI), with native plugins for Claude Code, Codex, Grok Build, and Antigravity. Narrative-architecture repair for fiction, venue-matched rules for professional prose. Based on StoryScope (arXiv:2604.03136).

2.5k
Imbad0202/
academic-research-skills

Academic Research Skills for Claude Code: research → write → review → revise → finalize

48k