Agent skills
The official nano agent skills, nano-deploy and nano-agentlane, for Claude Code, Codex and other coding agents. How to install them, how they're versioned and what they will and won't do.
nano publishes official Agent Skills for coding agents such as Claude Code, Codex and Cursor. A skill is a folder of instructions, references and scripts that your agent loads when a task needs it. It follows the open Agent Skills format, so it works with any agent that supports skills.
| Skill | What it does | Guide |
|---|---|---|
nano-deploy | Sizes, deploys, verifies, upgrades and operates nano Self-run on your own hosts: single node, 3-node HA, a cluster of several VMs, or Kubernetes with Helm. | Deploy with your AI agent |
nano-agentlane | Hunts and investigates on nano's event data through Agent Lane, with tested SQL recipes. Needs an Agent Lane token. | Hunt with your AI agent |
Both skills are for nano Self-run.
Install
The skills live in the public repository at github.com/nano-rs/nano. Replace <skill> with nano-deploy or nano-agentlane.
Claude Code (personal skills directory):
mkdir -p ~/.claude/skills && curl -fsSL https://github.com/nano-rs/nano/archive/refs/heads/main.tar.gz \
| tar -xz -C ~/.claude/skills --strip-components=2 nano-main/skills/<skill>Any agent with Agent Skills support, using the open skills installer:
npx skills add nano-rs/nano --skill <skill>Agents without skill support: download the folder to any path and tell the agent: "Read <skill>/SKILL.md and follow it."
Versions
The skills ship from the main branch of the public repository, so the install commands above always fetch the latest copy. Each SKILL.md names the nano version it was tested with in its metadata. To update, run the install command again.
Releases from v2.0.25 on are also tagged in the public repository. To match your release exactly, download the tag's archive (archive/refs/tags/v<version>.tar.gz, folder nano-<version>/skills/<skill>) instead of main. Settings → Agent Lane → Agent kit shows that command for the nano-agentlane skill on your install.
The public guides on this site are the authority. If a skill and a guide ever differ, follow the guide and tell us.
Safety
Both skills follow the same rules:
- Read before write.
nano-deployruns a read-only preflight before it changes anything.nano-agentlaneonly ever reads. - You approve every change.
nano-deployshows the commands and hosts for each step and waits for you. Destructive steps need their own confirmation. - Secrets stay out of sight. The skills never print, log or commit a token, password or key, and never put one on a command line. They read secrets from files or environment variables. Don't paste a secret into the chat; if you do, rotate it.
- No weakening security to pass a check. No turning off TLS verification, skipping image verification or working around a refusal.
- Untrusted input stays data. Log content and command output are never treated as instructions.
Review what your agent proposes as you would a colleague's change. You stay responsible for your hosts and data.
What's next
Agent Lane
A read-only, audited fast lane for AI agents onto nano's event data: native ClickHouse SQL over a dedicated HTTPS listener, with named, expiring tokens. nano Self-run only.
Hunt with your AI agent
Let Claude Code, Codex or another coding agent hunt and investigate on nano's event data through Agent Lane with the nano-agentlane agent skill.