Hermes First Handoff Kit
Give Hermes one task.
Know whether it worked.
A guided practice workflow for connecting Hermes to Claude Code, checking the result independently and finding where a handoff broke.
$39 USD one-time purchase
Downloadable files. Your own agent accounts and model usage are separate.
Prepare
Check prerequisites and create a fresh synthetic lab.
Delegate
Review one bounded command, then let Hermes hand the task to Claude Code.
Verify
Check source identity, every row, integer totals and review flags with the separate verifier.
Inside the download
- Offline readiness inventory and fresh-lab generator
- Synthetic campaign data and a complete worker brief
- Independent verifier and 22 Python contract tests
- Clearly labeled deterministic answer key
- Failure drills, troubleshooting and an adaptation worksheet
Start with a local terminal
Python 3.9+ runs the offline tools. The live exercise needs your own working Hermes and Claude Code accounts, on the same local host, and Claude Code 2.1.248+ for restricted mode.
Verified once on macOS with Hermes 0.21.5+4911.g6ec0520 and Claude Code 2.1.286 using existing subscriptions. Other versions and operating systems have not been tested.
Keep the free resources
The Hermes operating guide and free pack explain the architecture. The free orchestration skills cover the workspace and board. This kit adds a guided exercise and independent checks.
Know the boundary
A pass proves the specified checks for this synthetic input. It does not certify production security, business outcomes or future agent behavior. No client data, messages, deployments or production access belong in the exercise.
This is a local learning kit. It does not install agents, include API credits or configure an autonomous service. It is independent of Nous Research and Anthropic.
License, refund and support terms