amika-cli
Skill301 starsChanged 11 days agoApache-2.0View on GitHub
Drive the `amika` CLI to create, inspect, and operate Amika rigs, the organization's own hosts, and agent chat sessions non-interactively. Use when a task involves `amika rig`, `amika host`, `amika send`, `amika auth`, `amika secret`, `amika service`, `amika snapshot`, or `amika scp`, including listing or retiring bring-your-own-compute hosts and discovering and debugging services from inside an Amika rig.
- Reads credentials
In this repository11 other files it gives its agents.
- AGENTS.mdAGENTS.md
- js/AGENTS.mdAGENTS.md
- sandbox-image/AGENTS.mdAGENTS.md
- amika-annotationsSkill
- create-cli-release-commitSkill
- create-sdk-release-commitSkill
- distill-sessionSkill
- make-stacked-prsSkill
- naive-reader-reviewSkill
- new-specSkill
- release-commit-sandbox-imageSkill
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 4 are distinctive, and 7 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
7
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
Use amika to operate hosted Amika rigs, the organization's own hosts, and agent chat sessions without hanging on prompts or interactive …
- Distinctive: line 13
Read only the reference relevant to the task:
- Very specific: line 28
When $AMIKA_RIG_NAME is set, it means you are running inside a rig, and it has that rig's name.
- Very specific: line 43
Only if the command succeeds and the referenced service is absent should you inspect the repository's .amika/config.toml.
- Distinctive: line 48
If service listing fails, diagnose the CLI or authentication error.
- Distinctive: line 54
Follow these rules unless the user explicitly wants an interactive session:
- Very specific: line 56
Prefer -o json or -o json-pretty when the command supports it.
- Very specific: line 58
Do not run rig connect, rig open (alias rig code), or bare rig ssh in a tool call.
- Very specific: line 60
Use amika rig ssh <rig> <command> for non-interactive remote work.
- Very specific: line 61
Use amika send to send messages to an agent that either creates a new rig or lives inside an existing rig.
- Distinctive: line 63
Check structured results for per-item errors.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Using the amika CLI
Use amika to operate hosted Amika rigs, the organization's own hosts, and
agent chat sessions without hanging on prompts or interactive sessions.
Hosted rigs are the default target. Run amika <command> --help before relying on a flag in a reference
because the installed binary is authoritative.
Read only the reference relevant to the task:
- Rig lifecycle, hosts, agents, auth, and secrets
- Services, repository config, and snapshots
- SSH, editors, and file transfer
- Structured output and non-interactive behavior
This skill ships inside the amika binary, versioned with it. Install or
upgrade it with amika skill install --global (home directory) or
amika skill install --local (current repository), optionally limited with
--agents claude,codex,cursor. Rerun it after upgrading amika so the skill
matches the installed CLI.
Inside an Amika rig
When $AMIKA_RIG_NAME is set, it means you are running inside a rig,
and it has that rig's name. Otherwise, you are running outside a rig. Inside a
rig, _self names it wherever a command takes a rig.
For any task involving a service, URL, port, preview, or local server, inspect the rig's live services first:
amika service list --rig _self -o jsonFind the referenced service in that result and use its provisioned url and
ports. The live service record is authoritative for how to reach the service.
A service whose URL is - still exists; it may not have a public URL yet.
Only if the command succeeds and the referenced service is absent should you
inspect the repository's .amika/config.toml. Its [services.<name>] entry
declares the expected container port and URL scheme, but it does not prove that
the service was provisioned or reveal its current public URL.
If service listing fails, diagnose the CLI or authentication error. Do not treat a failed lookup as evidence that the service is absent and skip directly to the config file.
Safe automation
Follow these rules unless the user explicitly wants an interactive session:
- Prefer
-o jsonor-o json-prettywhen the command supports it. - Supply required confirmation flags up front.
- Do not run
rig connect,rig open(aliasrig code), or barerig sshin a tool call. They require a controllable terminal. - Use
amika rig ssh <rig> <command>for non-interactive remote work. - Use
amika sendto send messages to an agent that either creates a new rig or lives inside an existing rig. - Check structured results for per-item errors. Do not rely only on the process exit code.
See non-interactive behavior for output shapes, confirmation flags, and commands that reject JSON output.
Task routing
| Need | Start with |
| Identify the current rig | $AMIKA_RIG_NAME |
| Find a service URL or port inside a rig | amika service list --rig _self -o json |
| Create, list, start, stop, or delete rigs | amika rig --help |
| Run a coding agent | amika send --help |
| Run a command remotely | amika rig ssh <rig> <command> |
| Copy files | amika scp --help |
| Manage service exposure | amika service --help |
| Capture a base environment | amika snapshot --help |
| Fork from a snapshot | amika rig create --snapshot <name> |
| List the org's own hosts | amika host list |
| Run a rig on one of those hosts | amika rig create --host <hostname> |
| Authenticate in a headless environment | Set AMIKA_API_KEY |
Critical facts
- Rig creation auto-detects the current git repository and makes a clean clone. Uncommitted work is not copied.
snapshot list -o jsonreturns{"items": [...]}; most list commands return a bare array.--secret env:KEY=NAMErefers to a stored secret name, not a literal value.amika sendcan report an agent failure throughis_error; inspect the result even when the request completed..amika/config.tomlis the reviewable source for how a repository configures a rig. It is the default, but can be overridden, so it's not a substitute for checking a running rig's live state.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

