OpenClaw Spawn Prompt makes Claude Code always-on
This GitHub repo packages a prompt that tries to recreate OpenClaw behavior inside Claude Code, so you can keep an agent always on, reachable over Telegram, and able to open PRs or improve itself without any separate installation. The maker says it was iterated over the last few days to handle technical edge cases and will keep evolving as users fix issues through it.
The interesting part here is not a new model or a new app, but a distribution hack: it turns an agent workflow into something you can bootstrap with a single pasted prompt.
- –Lowest-friction angle: no install, no setup wizard, just Claude Code plus a prompt.
- –The pitch is persistence and autonomy, not raw capability.
- –Self-improving via PRs is a neat loop, but it also means the system’s quality will depend on prompt hygiene over time.
- –This will appeal most to power users already living in Claude Code, not to people looking for a polished end-user product.
DISCOVERED
49d ago
2026-04-08
PUBLISHED
49d ago
2026-04-08
RELEVANCE
AUTHOR
imusingreddityay