Templates

20 ready-to-run agent templates. Each is a standalone project that uses OpenSentinel as a library. Clone, add your API key, and run.

Overview

Templates are the fastest way to go from zero to a working AI agent. Each template is a standalone project under 200 lines that demonstrates a specific use case. They serve as both acquisition channels (discover OpenSentinel through a use case you care about) and starting points for custom agents.

  • 20 templates covering sales, recruiting, DevOps, trading, support, content, security, and more
  • Each template is a complete, runnable project with its own package.json and README
  • Uses opensentinel as an NPM dependency — all the power, none of the complexity
  • Clone, add your API key, and run in under 60 seconds
No database required. Templates use OpenSentinel in library mode by default. They need only a CLAUDE_API_KEY. Add a DATABASE_URL if you want persistent memory.

Quick Start

Pick any template and have it running in four commands:

Terminal
$ git clone https://github.com/dsiemon2/OpenSentinel.git
$ cd OpenSentinel/templates/ai-sales-agent
$ bun install
$ CLAUDE_API_KEY=sk-ant-... bun run start
That's it. The agent is now running. Replace ai-sales-agent with any template name from the list below.

Full Template List

All 20 templates with descriptions and key features:

TemplateDescriptionKey Features
ai-web-monitor Monitor web pages for changes with intelligent alerts URL tracking, diff detection, smart notifications
ai-sales-agent Lead research, scoring, and outreach drafting Company research, lead scoring, email drafts
ai-recruiter Candidate screening, ranking, and interview prep Resume parsing, skill matching, interview questions
ai-devops-agent Server monitoring, log analysis, incident response Log parsing, alerting, runbook execution
ai-trading-researcher Market research, sentiment scoring, trend detection News analysis, price tracking, sentiment scoring
ai-customer-support Ticket triage, response drafting, escalation Auto-categorize, draft replies, SLA tracking
ai-content-creator Multi-platform content from a single brief Blog posts, tweets, LinkedIn, email newsletters
ai-security-monitor Auth log analysis, network audit, file integrity Failed login detection, port scanning, hash checks
ai-code-reviewer PR review, bug detection, test coverage Line-by-line review, security flags, suggestions
ai-data-analyst Dataset profiling, insights, anomaly detection CSV/JSON parsing, statistical analysis, charts
ai-email-assistant Inbox triage, reply drafting, action extraction Priority sorting, draft replies, calendar parsing
ai-meeting-assistant Transcript summaries, action items, weekly digests Meeting notes, task extraction, follow-up tracking
ai-competitor-tracker Product monitoring, pricing changes, hiring signals Website monitoring, price alerts, job board tracking
ai-seo-optimizer Page scoring, meta optimization, content outlines SEO audit, keyword analysis, meta tag generation
ai-legal-reviewer Contract review, risk flagging, amendments Clause analysis, risk scoring, redline suggestions
ai-social-listener Brand monitoring, sentiment, trend detection Mention tracking, sentiment analysis, trend alerts
ai-documentation-writer Auto-generate API refs, guides, changelogs Code analysis, JSDoc parsing, markdown generation
ai-onboarding-agent Personalized onboarding plans, Q&A, tracking Custom checklists, progress tracking, FAQ answers
ai-inventory-manager Stock tracking, demand forecasting, PO generation Inventory levels, reorder alerts, purchase orders
ai-real-estate-analyst Property analysis, market research, ROI estimation Comps analysis, rental yield, market trends

Creating Your Own Template

Build a custom agent in four steps:

Copy an existing template

Start from the template closest to your use case:

Terminal
$ cp -r templates/ai-sales-agent templates/my-custom-agent
$ cd templates/my-custom-agent
Install OpenSentinel as a dependency
Terminal
$ bun add opensentinel
Configure and write your agent logic

Edit index.ts with your custom system prompt and tool usage:

index.ts
import { configure, chatWithTools } from 'opensentinel';

configure({
  CLAUDE_API_KEY: process.env.CLAUDE_API_KEY,
});

const systemPrompt = `You are a specialized AI agent for [your use case].
Your responsibilities:
1. [Task one]
2. [Task two]
3. [Task three]`;

const result = await chatWithTools([
  { role: 'system', content: systemPrompt },
  { role: 'user', content: 'Start your first task' }
]);

console.log(result.text);
Run your agent
Terminal
$ CLAUDE_API_KEY=sk-ant-... bun run start
💡
Use chatWithTools() for full power. This gives your agent access to all 123 built-in tools: shell execution, web browsing, file operations, search, code analysis, and more. The AI decides which tools to use based on the task.

Template Structure

Every template follows the same minimal structure:

File Tree
templates/ai-sales-agent/
index.ts          # Main agent logic (under 200 lines)
package.json      # Dependencies (opensentinel + any extras)
README.md         # Usage docs, configuration, examples

index.ts

The main file contains the agent's system prompt, configuration, and core loop. It imports configure and chatWithTools from opensentinel and defines the agent's behavior in a single file.

package.json

Minimal dependencies. Every template lists opensentinel as its primary dependency, plus any use-case-specific packages (e.g., a CSV parser for the data analyst template).

package.json
{
  "name": "ai-sales-agent",
  "version": "1.0.0",
  "scripts": {
    "start": "bun run index.ts"
  },
  "dependencies": {
    "opensentinel": "latest"
  }
}

README.md

Each template includes documentation covering what the agent does, how to configure it, example conversations, and ideas for customization.

ℹ️
Templates are starting points, not limits. Feel free to add persistent memory (DATABASE_URL), connect messaging channels (TELEGRAM_BOT_TOKEN), or add integrations. A template can grow into a full OpenSentinel deployment.