algochains-mcp-server 22.4.0

AlgoChains MCP Server — 482 tools for live algo trading: Tradovate/Alpaca/OANDA, Databento/Polygon data, Onyx RAG, Graphiti KG, AlphaLoop, HMM regimes, autonomous marketplace pipeline, Kalshi, real data only.

AlgoChains MCP Server

---

The only MCP server with live futures bots, real fill data, real-time ML inference, and ~482 tools across 20 domains — all backed by real APIs, zero synthetic data.

Connect your AI assistant (Claude, Cursor, ChatGPT) to your trading infrastructure in 3 commands. Ask Claude "What's my MNQ P&L today?" — it calls Tradovate, gets the real answer, and tells you.

You ask Claude:                    Claude calls:                    Server calls:
"What's my NQ position?"   →  get_positions()           →  Tradovate API → real data
"Run a backtest on MNQ"    →  run_backtest()             →  Databento tick archive
"Is the market trending?"  →  detect_market_regime()    →  Polygon + FRED → analysis
"Check my MNQ bot health"  →  get_bot_health(bot="MNQ") →  launchd + logs → live state

---

Quick Install

Option 1 — pipx (recommended, works on macOS Homebrew Python)

pipx install algochains-mcp-server
python scripts/quickstart.py --generate-config cursor
python scripts/quickstart.py --mode demo

Option 2 — pip in a virtual environment

python3 -m venv .venv && source .venv/bin/activate
pip install algochains-mcp-server

Option 3 — editable install from source (for development / contributors)

git clone https://github.com/AlgoChains/algochains-mcp-server.git
cd algochains-mcp-server
pip install -e ".[http,supabase,auth]"
python scripts/quickstart.py --mode demo

Option 4 — Homebrew (macOS)

brew tap algochains/algochains
brew install algochains

Homebrew Python (macOS): If you get a PEP 668 "externally-managed-environment" error, use pipx install algochains-mcp-server — pipx manages its own venv automatically. Do NOT use --break-system-packages.

That's it. Your AI now has 148 tools (smart mode) available immediately. Add broker credentials for live trading access. See Option C for live credentials.

---

Smart Mode vs Full Mode

AlgoChains exposes tools in two tiers, controlled by ALGOCHAINS_TOOL_MODE:

Mode	Tools Exposed	Token Cost	When to Use
Smart (default)	148 curated	~4K tokens	Cursor, Windsurf (80-tool limit), everyday use
Full (ALGOCHAINS_TOOL_MODE=full)	478 tools	~40K tokens	Claude Code, full agentic sessions

Smart mode includes: all live bot tools, market data, signals, research/backtest, Onyx RAG, prop fund pipeline, position sizing, broker management, and order execution. Everything you need 95% of the time.

Full mode unlocks the remaining 330 tools: advanced DeFi, Kalshi order placement, multi-tenant SaaS, QuantConnect integration, alt-data pipelines, and more.

discover_tools — Find Any Tool Without Full Mode

Even in smart mode, you can find and use any of the 478 tools:

# Ask the server to find the right tool for your task
discover_tools("walk-forward validation with leakage check")
# → Returns: walk_forward_test, run_mcpt_validation, analyze_overfitting

# Then call it
execute_dynamic_tool("walk_forward_test", {"symbol": "MNQ", "lookback_days": 252})

This provides 99.6% token reduction vs exposing all 478 tools (arXiv:2603.20313).

---

Tool Domains

All ~482 tools organized across 20 domains:

#	Domain	Smart	Full	Key Tools
1	Market Data	14	22	get_quote, get_ohlcv, get_tick_data, get_options_chain, get_footprint_chart, get_dark_pool_volume
2	Signals & Analysis	12	18	generate_signal, detect_market_regime, get_ensemble_vote, compute_gex, read_tape, pair_trade_signal
3	Research & Backtesting	10	16	run_backtest, walk_forward_test, run_mcpt_validation, validate_strategy_metrics, analyze_overfitting
4	Position Sizing	6	8	compute_r_multiple_size, compute_volatility_targeted_size, compute_idm, dual_size_conservative
5	Options Analytics	4	6	compute_option_greeks, find_optimal_strike, get_options_chain, unusual_options_activity
6	Prop Fund Pipeline	8	10	evaluate_strategy_for_prop_fund, simulate_prop_fund_evaluation, list_prop_funds, check_rithmic_status
7	Broker Management	6	15	check_all_broker_credentials, connect_broker, get_broker_onboarding_guide, store_api_key
8	Account Protection	6	8	check_protection_status, record_stop_event, lock_instrument, check_rate_limit_status
9	Order Execution	8	12	place_order, place_bracket_order, cancel_order, smart_route_order, execute_twap
10	Emergency / Destructive	3	5	flatten_all_positions, cancel_all_orders, emergency_stop, trip_circuit_breaker
11	Intelligence (Onyx + Macro)	10	14	onyx_ask, onyx_search, get_macro_signals, get_us_economic_indicators, get_fed_policy_signals
12	Prediction Markets	8	12	get_prediction_markets, search_prediction_markets, get_kalshi_settlements, place_kalshi_order
13	Skills Bridge	5	5	list_skills, get_skill_detail, search_skills, get_skills_for_task, invoke_moltbook_debate
14	Agent Memory	6	8	get_openclaw_memory, store_trade_lesson, get_current_regime, get_openclaw_state_summary
15	Live Bot Intelligence	12	18	get_bot_health, get_live_bot_metrics, get_bot_position_state, get_ai_pipeline_health, restart_trading_bot
16	Desktop Tower / Dispatch	4	8	dispatch_tower_job, get_tower_job_status, run_tower_backtest, sync_to_tower
17	Performance Reporting	4	6	generate_bot_tearsheet, get_bot_metrics_full, run_mcpt_pipeline, capture_learning_signal
18	Platform / SaaS	8	20	join_waitlist, create_support_ticket, track_platform_event, get_analytics_summary
19	AlphaLoop / Evolution	12	22	run_alphaloop_cycle, get_alphaloop_results, get_algochains_telos, send_ntfy_notification
20	Temporal Knowledge Graph (Graphiti)	2	4	graphiti_search, graphiti_temporal_query, graphiti_health, graphiti_add_episode — advisory agent_memory, never broker truth; fails closed graphiti_unavailable

Domain 20 (Graphiti): an advisory temporal context graph over Neo4j (getzep/graphiti), isolated in .venv-graphiti (Python 3.13) on the control-tower side. graphiti_search/graphiti_health are Tier-1 reads; graphiti_add_episode is WRITE_LOCAL (discover-only). It is agent_memory authority — never broker truth, never a trading dependency. discover_tools("graphiti") surfaces them. See control-tower docs/GRAPHITI_INTEGRATION_MEGAPROMPT.md.

---

Live Bot Showcase

AlgoChains runs 4 live futures bots on Tradovate. Their state, fills, ML pipeline health, and brackets stream through this MCP server in real time.

Bot	Symbol	Strategy	Live Since	Key MCP Tool
MNQ_Upgraded_Scalper	MNQ	7-AI ensemble, 5-min bars	Dec 2024	get_bot_health(bot="MNQ")
CL_Swing_Scalper	CL	FinBERT sentiment + momentum	Jan 2025	get_bot_health(bot="CL")
MES_EMA_Swing	MES	EMA pullback + regime detection	Feb 2025	get_bot_health(bot="MES")
NQ_EMA_Swing	NQ	Trend following + foundation model	Feb 2025	get_bot_health(bot="NQ")

get_bot_health — Full e2e Signal→Order→Fill Trace

# Returns: process state, position, bracket status, AI pipeline health,
#          ml_env_flags (MASSIVE_NEWS_FEATURES, MASSIVE_PCR_FEATURES, MASSIVE_HALT_GUARD),
#          cc_health (Command Center last-seen, WS status, Databento live feed age),
#          signal_health (params, risk_bootstrap, bot_version, trading_mode),
#          e2e_sentinel (signal→order→bracket→fill lifecycle state)
health = get_bot_health(bot="MNQ")

# All 4 bots in one call
status = get_all_bot_ops_status()
# Returns: process + position + bracket + pipeline snapshot for MNQ/CL/MES/NQ

No credentials needed if you have ALGOCHAINS_BRIDGE_API_KEY. Read-only.

---

Desktop Tower Dispatch

Heavy ML workloads (hyperparameter sweeps, walk-forward validation, feature importance) run on the desktop tower (teespc-1, 100.89.114.31) via dispatch_tower_job. The Mac stays clean.

# Dispatch a backtest or ML job to the GPU tower
dispatch_tower_job(
    job_type="backtest",
    params={"strategy": "mnq_scalper", "lookback_days": 252, "wfv_windows": 12}
)

# Check job status
get_tower_job_status(job_id="job_abc123")

From the CLI (ac command):

# Not yet in ac — see CLI_GAP_ANALYSIS.md for ac tower subcommand roadmap
python3 -c "
from algochains_mcp.algoclaw.desktop_tower import dispatch_tower_job
dispatch_tower_job('backtest', {'strategy': 'mnq_scalper', 'lookback_days': 90})
"

What runs where:

Component	MacBook (execution)	Desktop Tower (ML/GPU)
Live bots (MNQ/CL/MES/NQ)	✅ launchd	—
Token Guardian, Kalshi daemon	✅ launchd	—
Command Center (:3333)	✅ cloudflared tunnel	—
Onyx RAG (100.89.114.31:8085)	—	✅
GPU/ML: FinBERT, Kronos, vLLM	—	✅
Heavy backtests via dispatch_tower_job	sends job →	✅ executes

---

Security

Authentication Tiers

Scope	How to Authenticate	What's Allowed
Public / demo	No credentials	Market data, Onyx search, regime detection
Team	ALGOCHAINS_BRIDGE_API_KEY	Bot metrics, positions (read-only)
Owner	OWNER_API_TOKEN	Order execution, bot restart, emergency stop

Localhost-Only Services

The following services bind to 127.0.0.1 only and are never exposed publicly:

• MCP server HTTP bridge (port 8765 / stdio)
• Command Center dev server (port 3333) — external access via Cloudflare Access tunnel only
• Onyx RAG stack (tower port 8085) — accessible via Tailscale VPN only

Hard-Coded Safety Limits

These cannot be overridden by any AI agent:

Daily loss limit:      $500   (hard stop, all orders blocked until midnight)
Max drawdown:          15%    (circuit breaker trips at 15% peak-to-trough)
Human confirmation:    required for all orders above $10K notional
AI loop detection:     5 identical calls in 60s → 30-minute order block
VIX gate:             all trades blocked when VIX > 35

Full safety documentation: SAFETY_MODEL.md

OWNER_API_TOKEN — Mutation Gating

Tools in danger tier 2 (order execution) and tier 3 (destructive) require OWNER_API_TOKEN in the request header. The HTTP bridge verifies this before dispatching. AI agents that do not supply it get a policy_denied error — not a soft warning.

# Set in .env (never commit)
OWNER_API_TOKEN=your-owner-token-here

---

What's New in v22.x

v22.4 (2026-04-06) — UX & Team Onboarding

• Complete README rewrite (plain English, team access)
• scripts/quickstart.py — interactive setup wizard with health checks
• SAFETY_MODEL.md — answers "is this safe?" for every failure mode
• tool_danger_tiers.py — machine-readable danger classification (0–3) for all 350+ tools
• HTTP bridge /tools endpoint now returns danger_tier, safe_in_demo_mode, etc.

v22.2 (2026-04-21) — Kalshi Pipeline + Model Integrity

• Kalshi prediction markets — AI ensemble → Kelly sizing → order execution
• Subscriber tools — JWT tier auth, get_subscriber_portfolio, get_marketplace_listings
• Unified path resolver (paths.py) — default_control_tower() works on Mac + WSL tower
• Data backend chain — Databento → Massive S3 (back to 2003) → Polygon → yfinance
• SHA-256 model integrity — startup check raises on tampered .pkl, XGBoost JSON companion, model_manifest.json
• Drawdown Triple Penance — drawdown_start_ts auto-logged on first daily loss hit (Bailey & LdP 2015)

v22.0 (2026-04-05) — MCP 2025-11-25 Full Compliance

• Elicitation (human confirmation for high-value trades)
• Durable Tasks (background backtest/optimization jobs)
• SSE streaming transport
• OIDC discovery endpoint
• Trading guardrails with circuit breakers
• AlphaLoop evolution daemon

See the full CHANGELOG.md for v23.x, v24.x, v25.x, and v26.x entries.

---

Quick Setup Options

Option A — Demo Mode (No Credentials, 1 Minute)

git clone https://github.com/AlgoChains/algochains-mcp-server.git
cd algochains-mcp-server
pip install -e ".[http,supabase,auth]"
python scripts/quickstart.py --mode demo

Available immediately (no credentials):

• get_quote("AAPL") — live price for any symbol
• detect_market_regime() — trending / ranging / choppy
• get_macro_signals() — macro environment analysis
• discover_tools() — find any of the 478 tools
• onyx_ask("any question") — knowledge base search

Option B — Paper Mode (Alpaca Paper, Free)

export ALPACA_API_KEY=your-paper-key
export ALPACA_SECRET_KEY=your-paper-secret
export ALPACA_PAPER=true
python scripts/quickstart.py --mode paper

Option C — Full Live Setup

cp .env.example .env
# Edit .env with Tradovate, Polygon, Databento, Slack credentials
python scripts/quickstart.py --health-check --mode live

Generate IDE Config

python scripts/quickstart.py --generate-config cursor         # Cursor
python scripts/quickstart.py --generate-config claude-desktop # Claude Desktop
python scripts/quickstart.py --generate-config windsurf       # Windsurf

---

Data Backends

AlgoChains uses a priority chain — best available source wins automatically:

Priority	Backend	Coverage	Use Case
1	Databento	XNAS.ITCH + XNYS.PILLAR; OHLCV-1d + OHLCV-1m	Futures tick data, live streaming
2	Massive S3	us_stocks_sip/day_aggs_v1/ back to 2003	Historical equity backtests, survival-bias-free universe
3	Polygon	REST bars + news snapshots	News features, intraday bars
4	yfinance	Free, ~5yr history	Dev fallback, swing bots

Force a specific backend: DATA_BACKEND=databento|massive|polygon|yfinance in .env.

---

Command Center

URL	Status	Notes
https://cc.algochains.io	Live	Cloudflare Access — authenticate with [e-mail]
http://localhost:3333	Local dev	Always accessible without auth

Run locally:

cd algochains-command-center
npm run dev   # starts on :3333

Start Cloudflare tunnel:

cloudflared tunnel run def269f2-6c52-471a-9648-c2fe631bc9bf >> logs/cloudflared_cc.log 2>&1 &

Dashboard panels (V22):

• Bot Status Cards — process state, uptime, last signal, AI confidence
• P&L Chart + Positions Table + Risk Dashboard
• Bracket Status Panel + AI Ensemble Health + Live Trade Validation Feed (SSE)
• Subscriber Protection Panel + System Health

---

Agentic Quick-Start Prompts

Copy these directly into Claude or Cursor:

Morning brief:
"Run get_macro_signals and get_live_bot_metrics. Summarize market conditions and P&L."

Bot health check:
"Run get_bot_health for all 4 bots. Flag anything that needs attention."

Pre-trade regime check:
"Before I place any orders, run detect_market_regime and check VIX. Should I trade today?"

Validate a backtest:
"Run validate_strategy_metrics: Sharpe 2.4, MaxDD 9%, WinRate 58%, 180 trades.
 Does it pass the MCPT gate? What's the DSR?"

Prop fund compatibility:
"Use evaluate_strategy_for_prop_fund: MNQ scalper, $600 max daily loss, $2500 max DD,
 $120 avg daily profit, holds overnight. Which fund should I target?"

Emergency system check:
"Run check_all_broker_credentials and check_rithmic_status. What's ready, what's missing?"

Tower dispatch:
"Dispatch an overnight Optuna sweep for MNQ to the desktop tower. 200 trials, Sharpe objective."

---

Supported Brokers

Broker	Asset Classes	Status
Tradovate	Futures (MNQ, CL, MES, NQ, ES, GC)	✅ Live
Alpaca	Equities, ETFs, Options, Crypto	✅ Live + Paper
OANDA	Forex (50+ pairs)	✅ Live
Interactive Brokers	Stocks, Futures, Options, Forex	✅ Live (ib_async)
Kalshi	Prediction markets (US events)	✅ Live
E*TRADE	Equities, Options, ETFs	✅ OAuth 1.0a
Rithmic	Futures via prop fund platforms	⏳ DRY_RUN (vendor NDA pending)
Charles Schwab	Equities, Options, Futures	⚠️ Stubs (OAuth 2.0 PKCE)

# Check all broker credential status at once
check_all_broker_credentials()   # masked — never exposes values

---

Architecture

Your AI (Claude / Cursor / ChatGPT)
         │
         │ MCP 2025-11-25 (stdio or HTTP + SSE)
         ▼
AlgoChains MCP Server
  ├── 478 tools / 148 smart-mode (19 domains)
  ├── Trading Guardrails (hard-coded limits, AI loop detection)
  ├── Account Protection (12 pre-trade guards)
  ├── Onyx RAG (semantic search — 400+ docs + 472 skills)
  └── Circuit Breakers (per-tool rate limits, daily loss stops)
         │
         ├── Tradovate     (MNQ, CL, MES, NQ futures — live fills)
         ├── Alpaca        (equities, crypto, options)
         ├── OANDA         (forex)
         ├── Databento     (tick-level data — XNAS.ITCH)
         ├── Massive S3    (day bars back to 2003)
         ├── Polygon       (real-time bars, news)
         └── FRED, CBOE, Kalshi, Polymarket  (macro / alt data)

Data policy: No synthetic data. No mock fills. No placeholder values. Every tool connects to a real API or fails closed with an explicit error.

---

Docs

File	Purpose
SAFETY_MODEL.md	Is this safe? Failure modes, guardrails, team access
CHANGELOG.md	Full version history
docs/GOTCHAS_AND_BUGS.md	Confirmed bugs, gotchas, operational surprises
docs/TRADOVATE_PARITY.md	Tradovate endpoint mapping vs community server
docs/CLI_GAP_ANALYSIS.md	ac CLI current commands + 10 missing subcommands roadmap
LATENCY_GUIDE.md	Measured tool call latencies (Mac M3 Max, real calls)
MARKETPLACE_CREATOR_GUIDE.md	Submit a validated bot to the marketplace
algoclaw/README.md	AlgoClaw agent skill system

---

Built by Tyler Reynolds — experimental AI trading infrastructure.

Safety · Changelog · Command Center · Marketplace

Experimental software connected to live trading accounts. Use at your own risk.