General

I research anything from Telegram now

I needed to compare three headless CMS platforms for a client project. Normally that means opening 15 tabs, reading marketing pages that all sound the same, checking Reddit for real opinions, and then trying to make sense of it all. Two hours minimum.

I realized I could just ask my agent. It searches the web, reads the pages, and gives me a straight answer. No tabs, no skimming, no context-switching between browser and notes.

The agent

The Basic Agent is a general-purpose AI running on your own server. No pre-installed skills, no locked-in workflow. It searches the web, runs code, analyzes files, and answers questions. Basically a research assistant that lives in Telegram and never sleeps.

Setting it up

About 7 minutes from start to first message:

Find the agent

Head to the Hiregents marketplace and find Basic Agent. Read the skill file if you want to know exactly what it can do. Everything is transparent.

Sign in

Click Deploy. You'll get a magic link to your email. No passwords, no account setup. Click the link and you're in.

Pick a plan

Hobby ($39/mo) or Pro ($49/mo). Both get you a private server. The difference is the AI model: Hobby runs Gemini Flash, Pro runs Claude Sonnet. Annual billing saves 25%.

Connect your keys

Two things needed: an OpenRouter API key (takes 30 seconds to create at openrouter.ai) and a Telegram bot token (message @BotFather on Telegram, type /newbot, follow the prompts). Both get validated in real-time before you can continue.

Review and deploy

Check the summary: agent name, plan, email. Hit Deploy. You'll go through Stripe checkout and land on a status page.

Wait about 5 minutes

The status page shows a progress circle while your server boots. Docker pulls, security gets configured, the agent installs its tools. You don't need to stay on the page. It'll be ready when you come back.

Open Telegram

Find your bot on Telegram (the one you created with BotFather). Send /start. Basic Agent is live.

What the conversation looks like

12:34

Basic Agent

online

Today

Compare Sanity, Contentful, and Strapi for a blog with 50k monthly readers. I care about pricing, speed, and DX.

12:34

Basic Agent

Searched and compared all three. Here's the breakdown: Sanity Pricing: Free tier generous (100K API calls/mo). Paid starts at $99/mo. Speed: GROQ query language is fast. CDN-backed. DX: Excellent. Fully customizable studio, TypeScript-first. Contentful Pricing: Free tier tight (25K records). Paid at $300/mo. Speed: REST + GraphQL. CDN delivery is solid. DX: Good SDK, but the web app is slow. Content modeling is rigid. Strapi Pricing: Free (self-hosted). Cloud starts at $29/mo. Speed: Depends on your hosting. No built-in CDN. DX: Great if you like owning the stack. Plugin ecosystem is hit-or-miss. For 50K readers on a blog: Sanity. Best free tier, fastest queries, and the studio is a pleasure to use. Strapi if you want full control and have DevOps capacity.

12:34

What are people saying about Sanity on Reddit? Any common complaints?

12:34

Basic Agent

Searched Reddit threads from the last 6 months: Common praise: Studio customization is the best in class GROQ is genuinely faster than GraphQL for content queries Free tier is real, not a bait-and-switch Common complaints: Pricing jumps sharply after free tier (some report surprise bills) Image pipeline can be slow without their CDN add-on Documentation has gaps for advanced use cases Vendor lock-in concern (GROQ is proprietary) Overall sentiment: positive, but people recommend watching your API usage closely on the paid tier.

12:34

Message

What I got

A full comparison with real user feedback in about 3 minutes. Official docs, pricing pages, Reddit threads, all pulled together. I sent my recommendation to the client that same afternoon.

Try it yourself

Deploy Basic Agent View agent profile