Tutorials
Learn to put agents to work
Step-by-step guides to the tasks worth doing well — how to benchmark harness and model combinations on your own work, connect agents to the channels your team already uses, and bring your own Claude or ChatGPT subscription so cloud agents run at zero model cost.
Bring your own subscription
Run cloud agents on the Claude or ChatGPT plan you already pay for
Connect your Claude Pro/Max subscription or ChatGPT plan once and eligible agents run on it at $0 model usage — no per-token metering, no second bill. Setup takes two minutes in Settings or one CLI command.
5 min read · Updated August 2026
Read the tutorialEvals & benchmarking
How to benchmark your skills in 2026
A practical playbook for benchmarking agent skills — harness × model combinations — on your real work instead of synthetic leaderboards. Build the matrix, hold conditions identical, and pick a winner in an afternoon.
8 min read · Updated July 2026
Read the tutorialBuzz
Run your Buzz agent in the cloud
Clone the agent you already run locally into AgentSky, then point buzz-acp at it. Always-on, nothing to host, memory that survives every session rotation. Takes about 15 seconds.
4 min read · Updated August 2026
Read the tutorialSlack
Your agents in Slack
Every agent gets its own Slack channel. Add an agent in one click, @mention it, and bring your teammates along — no commands to learn, no switching apps.
3 min read · Updated August 2026
Read the tutorialTelegram
Your agents in Telegram
One bot, all your agents. Connect once, see who's active, and switch with a tap — every agent you create shows up in the same chat automatically.
3 min read · Updated August 2026
Read the tutorial