Summary

Adds --openclaw flag to kalibr prompt CLI command, outputting a fully autonomous onboarding prompt for OpenClaw / Hermes agent orchestrators.

Changes

kalibr/cli/prompt_cmd.py

• Added --openclaw boolean flag via typer.Option
• Added OPENCLAW_PROMPT_TEXT — a step-by-step autonomous setup sequence covering:
  1. pip install kalibr huggingface_hub
  2. kalibr auth --agent --email <user-email> (asks user for email)
  3. Credential sourcing into ~/.zshrc / ~/.bashrc
  4. openclaw plugin install kalibr && openclaw plugin configure kalibr
  5. ~/clawd/ workspace scaffold (SOUL.md, USER.md, HEARTBEAT.md, AGENTS.md)
  6. ~/clawd/kalibr-integration/ with pipeline files from examples/
  7. kalibr init && kalibr verify
  8. Smoke test: classify_and_route()
  9. Done report with dashboard link + claim URL
• Clipboard copy and label both updated to reflect --openclaw mode

README.md

• New section "## OpenClaw / Hermes Agent Setup" inserted immediately after "## Coding Agent Setup"
• Documents kalibr prompt --openclaw, the 9-step flow, and usage examples

Usage

# Standard (Claude Code / Cursor)
kalibr prompt

# Full autonomous OpenClaw / Hermes onboarding
kalibr prompt --openclaw

Testing

• kalibr prompt — unchanged behavior, outputs PROMPT_TEXT to stdout and clipboard
• kalibr prompt --openclaw — outputs OPENCLAW_PROMPT_TEXT, clipboard label updated