Your X account grows while you sleep.
Research, draft in your voice, post with approval, analyze
$ npx seclaw add content-agent --key YOUR_TOKENWhat this looks like in Telegram
seclaw
3 draft variants ready: #1 Educational: "Most people think RAG is about embeddings. It's actually about chunking. Here's what I learned building a production system..." #2 Personal story: "I spent 3 months building the wrong RAG system. The fix took 2 hours. Thread" #3 Contrarian: "Hot take: RAG is already obsolete. Context windows are eating retrieval for lunch." Reply 1, 2, or 3 to post.
— Content Agent
You
Go with #2
seclaw
Posted to X! Tracking engagement — I'll report back in 24h with impressions, likes, and replies.
— Content Agent
Research: scans trends, competitors, and past performance
Drafting: creates 3 draft variants per content idea
Posts approved content to X via API
Research (09:00) --> Draft (11:00) --> Preview (Telegram) --> [Approve] --> Post
--> [Reject] --> SkipVoice matching
Reads your style guide, mimics your tone
3 drafts always
Educational, story/personal, contrarian
Human-in-the-loop
Requires Telegram approval for every post
Research-driven
Morning trend scan informs drafts
File-based memory
Analytics and style guide persist across restarts
OpenClaw proved the demand — 68K+ GitHub stars. But it ships with zero container isolation: API keys exposed to every MCP container, your home directory mounted to root-privileged processes, port 5678 open with no auth.
Every seclaw template runs in Docker with hard guardrails — no root, no host access, no exposed ports. Security enforced at runtime, not in the system prompt.
Purchase at seclawai.com/templates/content-agent
Run npx seclaw add content-agent --key YOUR_TOKEN
Create your style guide at /shared/voice/style-guide.md
Message your bot on Telegram to get started
$49 one-time. No subscription. Self-hosted.
npx seclaw add content-agent --key YOUR_TOKEN