Use Telegram to run a VC research agent
Jonathan Shriftman wired OpenClaw to Telegram so he could ask it to research companies and help draft internal tools.
- What they did
- He installed OpenClaw on a Hostinger VPS priced at $6.99 per month, added a Claude API key, and linked the agent to Telegram through BotFather. He named it Pao and started by asking it to inspect its own security. He then used chat commands to have it code and launch a CRM replacement built with Supabase and Vercel.
- How it went
- The initial setup took about 45 minutes. The security review cost $0.30, and the first CRM version ran to about $25. Claude repeatedly ran into limits, deployments became stuck, and he said the sub-agents were opaque; a portfolio-monitoring feature has not been built yet.
- Tips
- Costs are usage-based, so a small experiment can grow quickly; one partner in the same work spent $150 in a week. This approach is best for people who already know how to handle API keys, SSH, and terminal work.
Try it yourself with OpenClaw
You are my analyst inside [Telegram]. For any [company name] or [URL] I send, investigate it and return a brief memo with [business model, founders or team, funding, rivals, and possible concerns]. Limit the memo to [400] words. Use sources for factual claims. If you need to pay for data, contact someone outside this chat, delete or change files, or submit anything anywhere, stop first and ask for my approval.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Use an agent to manage roadmap, support, calendar, and tasks
A developer set up OpenClaw to sort support mail, update a roadmap, handle calendar items, and keep track of todos and sponsorships.
Set up OpenClaw bots for a repo, notes, and Slack
They used OpenClaw to keep a personal repository updated, capture notes, and handle Slack summaries and questions.
Use a desktop agent to move files with checks
A user drove a desktop agent through file moves in real apps, and learned that checkpoints matter when the interface lies about success.
Route each task to the cheapest capable model
A user changed OpenClaw so it could choose lower-cost models for CRM work and email drafting, with manual approval before any outside action.
Gather supplier quotes by email
An agent contacted shower glass suppliers, pulled together replies, and kept the email thread moving.
Automate inbox checks, research digests, and site audits
Jenny Ouyang uses two OpenClaw agents with scheduled jobs to scan email, draft research digests, and monitor a website.