{"id":46082628,"url":"https://github.com/augiemazza/varrd","last_synced_at":"2026-03-04T05:01:00.783Z","repository":{"id":341396220,"uuid":"1169968122","full_name":"augiemazza/varrd","owner":"augiemazza","description":"  From idea to statistically validated trade setup. VARRD is more than a back tester you can query with natural language — it's your home for safely researching market edges without the invisible statistical mistakes that invalidate most backtests. Let your agent autonomously discover edges  with peace of mind. ","archived":false,"fork":false,"pushed_at":"2026-03-01T16:51:44.000Z","size":35,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-03-01T17:40:39.145Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/augiemazza.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2026-03-01T14:03:47.000Z","updated_at":"2026-03-01T16:34:07.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/augiemazza/varrd","commit_stats":null,"previous_names":["augiemazza/varrd"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/augiemazza/varrd","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/augiemazza%2Fvarrd","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/augiemazza%2Fvarrd/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/augiemazza%2Fvarrd/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/augiemazza%2Fvarrd/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/augiemazza","download_url":"https://codeload.github.com/augiemazza/varrd/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/augiemazza%2Fvarrd/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":30071895,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-03-04T03:25:38.285Z","status":"ssl_error","status_checked_at":"2026-03-04T03:25:05.086Z","response_time":59,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-03-01T16:05:19.904Z","updated_at":"2026-03-04T05:01:00.749Z","avatar_url":"https://github.com/augiemazza.png","language":"Python","funding_links":[],"categories":["Trading \u0026 Backtesting","Community Projects","🔥 AI Powered Systematic Trading Systems"],"sub_categories":["Trading \u0026 Finance"],"readme":"# VARRD — Trading Edge Discovery\n\n[![PyPI](https://img.shields.io/pypi/v/varrd)](https://pypi.org/project/varrd/)\n[![MCP](https://img.shields.io/badge/MCP-7_tools-blue)](https://app.varrd.com/mcp)\n[![Transport](https://img.shields.io/badge/transport-Streamable_HTTP-green)](https://app.varrd.com/mcp)\n[![License](https://img.shields.io/badge/license-MIT-lightgrey)](LICENSE)\n\n**Institutional-grade quant research. Describe any trading idea in plain English, get statistically validated results with exact trade levels.**\n\nAny AI can backtest a strategy. VARRD guarantees it was done right — with K-tracking, Bonferroni correction, OOS lock, lookahead detection, and 4 other integrity guardrails enforced at infrastructure level.\n\n---\n\n## MCP Server — 7 Tools, 4 Prompts\n\n**Endpoint:** `https://app.varrd.com/mcp`\n**Transport:** Streamable HTTP (2025-03-26 spec)\n**Auth:** Anonymous access with auto-provisioned credits. No API key required.\n\n```json\n{\n  \"mcpServers\": {\n    \"varrd\": {\n      \"transport\": {\n        \"type\": \"streamable-http\",\n        \"url\": \"https://app.varrd.com/mcp\"\n      }\n    }\n  }\n}\n```\n\nWorks with **Claude Desktop**, **Claude Code**, **Cursor**, and any MCP-compatible client.\n\n### MCP Tools\n\n| Tool | Cost | What It Does |\n|------|------|--------------|\n| `research` | ~$0.25 | Multi-turn quant research with VARRD AI. Orchestrates 15 internal tools (data loaders, charting, event studies, backtests, optimization). Follow `context.next_actions` each turn. |\n| `autonomous_research` | ~$0.25 | AI discovers edges for you. Give it a topic, it generates hypotheses from its market knowledge base, runs the full pipeline, returns validated results. |\n| `scan` | Free | Scan saved strategies against live market data. Returns exact dollar entry, stop-loss, and take-profit prices for every active signal. |\n| `search` | Free | Find saved strategies by keyword or natural language. Returns matches ranked by relevance with win rate, Sharpe, edge status. |\n| `get_hypothesis` | Free | Full details on any strategy: formula, entry/exit rules, win rate, Sharpe, profit factor, max drawdown, version history. |\n| `check_balance` | Free | View credit balance and available credit packs. |\n| `reset_session` | Free | Kill a broken research session and start fresh. |\n\n### MCP Prompts\n\n| Prompt | Description |\n|--------|-------------|\n| `test-trading-idea` | Test any trading idea with real market data and statistical validation |\n| `whats-firing-now` | Scan your validated strategies and show what's actively firing |\n| `discover-edges` | Let VARRD's autonomous AI discover trading edges on a topic |\n| `find-strategies` | Search your strategy library by keyword or concept |\n\n---\n\n## Quick Start — CLI\n\n```bash\npip install varrd\n\n# Research an idea (auto-follows the full workflow)\nvarrd research \"When wheat drops 3 days in a row, is there a snap-back?\"\n\n# What's firing right now?\nvarrd scan --only-firing\n\n# Search your saved strategies\nvarrd search \"momentum on grains\"\n\n# Let VARRD discover edges autonomously\nvarrd discover \"mean reversion on futures\"\n\n# Check credits\nvarrd balance\n```\n\n## Quick Start — Python\n\n```python\nfrom varrd import VARRD\n\nv = VARRD()  # auto-creates free account\n\n# What's firing right now?\nsignals = v.scan(only_firing=True)\nfor s in signals.results:\n    print(f\"{s.name}: {s.direction} {s.market} @ ${s.entry_price}\")\n\n# Research a trading idea\nr = v.research(\"When RSI drops below 25 on ES, is there a bounce?\")\nr = v.research(\"test it\", session_id=r.session_id)\nprint(r.context.has_edge)       # True / False\nprint(r.context.edge_verdict)   # \"STRONG EDGE\" / \"NO EDGE\" / etc.\n\n# Get the trade setup\nr = v.research(\"show me the trade setup\", session_id=r.session_id)\n```\n\n---\n\n## Authentication \u0026 Passkey\n\nFirst use auto-creates an account. You'll receive a **passkey** (`VARRD-XXXXXXXXXXXXXXXX`) saved to `~/.varrd/credentials`.\n\n```\n  VARRD account created.\n  Your passkey: VARRD-A3X9K2B7T4M8P1Q6\n  Saved to: ~/.varrd/credentials\n```\n\n**Link to browser:** Go to [app.varrd.com](https://app.varrd.com), click \"Link your AI agent\", enter your passkey with an email and password. Your agent's strategies and credits merge into your account.\n\n```python\nv = VARRD(api_key=\"your-key\")         # Python\n```\n```bash\nvarrd --key your-key scan              # CLI\nexport VARRD_API_KEY=your-key          # Environment variable\n```\n\n---\n\n## What You Get Back\n\n### Edge Found\n```\nSTRONG EDGE: Statistically significant vs both zero and market baseline.\nThe pattern produces real returns that beat market drift.\n\n  Direction: LONG\n  Win Rate:  62%\n  Sharpe:    1.45\n  Trades:    247\n  K:         3 (tests run on this hypothesis)\n\n  Trade Setup:\n    Entry:       $5,150.25\n    Stop Loss:   $5,122.00\n    Take Profit: $5,192.50\n    Risk/Reward: 1.5:1\n```\n\n### No Edge\n```\nNO EDGE: Neither test passed — no tradeable signal found.\n\nThis is a valid result. You found out for 25 cents\ninstead of $25,000 in live losses.\n```\n\n---\n\n## The Research Flow\n\n```\nYour idea (plain English)\n    |\n    v\n  Chart pattern — see the actual signals on real price data\n    |\n    v\n  You approve — sanity check before spending statistical power\n    |\n    v\n  Statistical test — event study or backtest with proper controls\n    |                 (K increments, fingerprints logged)\n    v\n  Edge verdict:  STRONG EDGE — beats zero AND beats market\n                 MARGINAL    — beats zero, doesn't clearly beat market\n                 NO EDGE     — no signal found\n    |\n    v\n  Trade setup — exact dollar entry, stop-loss, take-profit\n```\n\nA typical session is 3-5 turns and costs ~$0.25.\n\n## 8 Statistical Guardrails (Infrastructure-Enforced)\n\n| Guardrail | What It Does |\n|-----------|--------------|\n| **K-Tracking** | Counts every test. 50 variations = higher significance bar. |\n| **Bonferroni Correction** | Multiple comparison penalty, automatic. |\n| **OOS Lock** | Out-of-sample is sacred. One shot. Locked forever. |\n| **Lookahead Detection** | Catches formulas that accidentally use future data. |\n| **Tools Calculate, AI Interprets** | AI never fabricates a number. Every stat from real data. |\n| **Chart \u003e Approve \u003e Test** | Must see and approve pattern before testing. |\n| **Fingerprint Deduplication** | Can't retest same formula/market/horizon twice. |\n| **No Post-OOS Optimization** | Parameters locked after OOS validates. |\n\n## Data Coverage\n\n| Asset Class | Markets | Timeframes |\n|------------|---------|------------|\n| **Futures** (CME) | ES, NQ, CL, GC, SI, ZW, ZC, ZS, ZB, TY, HG, NG + 20 more | 1h and above |\n| **Stocks/ETFs** | Any US equity | Daily |\n| **Crypto** (Binance) | BTC, ETH, SOL + more | 10min and above |\n\n15,000+ instruments total.\n\n## Pricing\n\n- **$2 free** on signup — enough for 6-8 full research sessions\n- **Research:** ~$0.20-0.30 per complete workflow\n- **ELROND council** (8 expert investigators): ~$0.40-0.60\n- **Multi-market** (3+ markets): up to ~$1\n- **Free tools:** scan, search, get_hypothesis, check_balance, reset_session\n- **Credit packs:** $5 / $20 / $50 via Stripe\n- Credits never expire\n\n---\n\n## Examples\n\nSee [`examples/`](examples/):\n\n- [`quick_start.py`](examples/quick_start.py) — 5 lines to get started\n- [`scan_portfolio.py`](examples/scan_portfolio.py) — Scan all strategies, show what's firing\n- [`research_idea.py`](examples/research_idea.py) — Full multi-turn research workflow\n- [`multi_idea_loop.py`](examples/multi_idea_loop.py) — Test many ideas in a loop\n- [`mcp_config.json`](examples/mcp_config.json) — MCP config for Claude Desktop\n\n## For AI Agents\n\nSee [`AGENTS.md`](AGENTS.md) for a structured guide with complete tool reference, response formats, and integration patterns.\n\n---\n\n## Links\n\n- **Web app:** [app.varrd.com](https://app.varrd.com)\n- **Website:** [varrd.com](https://www.varrd.com)\n- **MCP endpoint:** `https://app.varrd.com/mcp`\n- **MCP Registry:** [registry.modelcontextprotocol.io](https://registry.modelcontextprotocol.io)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faugiemazza%2Fvarrd","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faugiemazza%2Fvarrd","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faugiemazza%2Fvarrd/lists"}