Sentinel AI

Autonomous financial intelligence platform for deep market research.

Sentinel AI is a CLI-based agent that conducts real-time financial research using live market data, SEC filings, web intelligence, and multi-provider LLM reasoning. Ask a question, and Sentinel autonomously gathers data, analyzes fundamentals, and delivers concise, data-backed answers.

Features

• Multi-Provider LLM Support — OpenAI, Anthropic, Google Gemini, xAI (Grok), OpenRouter, Ollama (local), Moonshot, DeepSeek
• Financial Data Engine — Real-time prices, historical data, financial metrics, SEC filings (10-K, 10-Q, 8-K), insider trades, institutional holdings
• Intelligent Web Research — Exa, Perplexity, and Tavily search integration with full-page reading via Playwright
• Persistent Memory — Semantic search over past conversations; remembers your preferences and research context
• Extensible Skills — SKILL.md-based workflows for complex analyses (e.g., DCF valuation models)
• WhatsApp Gateway — Connect Sentinel to WhatsApp for mobile financial research
• Scheduled Monitoring — Heartbeat system for periodic market checks and alerts
• Rich Terminal UI — Interactive CLI with formatted tables, syntax highlighting, and real-time streaming

---

Architecture

src/
├── agent/          # Core reasoning loop, system prompts, context management
├── tools/          # Financial search, web research, browser, filesystem
├── components/     # Terminal UI widgets (Ink/pi-tui)
├── commands/       # Slash command definitions
├── memory/         # Persistent semantic memory system
├── skills/         # Extensible SKILL.md workflows
├── gateway/        # WhatsApp messaging integration
├── cron/           # Scheduled job execution
├── model/          # Multi-provider LLM abstraction
└── utils/          # Configuration, caching, token management

---

Architecture Diagram

---

---

data flow Diagram

---

---

Quick Start

Prerequisites

• Bun runtime (v1.0+)
• At least one LLM API key (OpenAI recommended)

Installation

# Clone the repository
git clone https://github.com/<your-username>/sentinel-ai.git
cd sentinel-ai

# Install dependencies
bun install

# Configure environment
cp env.example .env
# Edit .env with your API keys

# Launch Sentinel
bun run start

Usage

# Start interactive session
bun run start

# Development mode (auto-reload)
bun run dev

# WhatsApp gateway
bun run gateway

Commands

Command	Description
/model	Switch LLM provider and model
/rules	View your custom research rules
/memory	Show what Sentinel remembers
/history	View recent conversation summaries
/heartbeat	Configure market monitoring
/clear	Clear the conversation
/help	Show keyboard shortcuts

Configuration

Environment Variables

Copy env.example to .env and configure:

• LLM Keys: OPENAI_API_KEY, ANTHROPIC_API_KEY, GOOGLE_API_KEY, etc.
• Financial Data: FINANCIAL_DATASETS_API_KEY
• Web Search: EXASEARCH_API_KEY, PERPLEXITY_API_KEY, or TAVILY_API_KEY
• Tracing: LANGSMITH_API_KEY for LangSmith observability

Settings

Persistent settings are stored in .sentinel/settings.json (auto-created). Use /model to switch providers interactively.

---

Development

# Type-check
bun run typecheck

# Run tests
bun test

# Watch mode
bun run dev

License

MIT — Copyright (c) 2026 Rajkaran Yadav