Setup
This skill should be used when the user asks to "set up getlark", "install larkci", "install the getlark CLI", "configure GETLARK_API_KEY", "authenticate with getlark", or runs `/getlark:setup`. Walks the user through installing @getlark/cli and authenticating via `getlark login` or by exporting GETLARK_API_KEY in their shell rc. One-time per machine — skip if `getlark --version` already works and the user is authenticated; defer to the action skills (`create-workflow`, `invoke-workflow`, `manage`, `validate-branch`) in that case.From its SKILL.md
npx -y skills add getlark/skills --skill setupAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its file declares
Copied from the file, not written here
The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
4.0 KB, as published. Nobody here has run it
setup
Install the getlark CLI globally via npm and authenticate so subsequent getlark commands work without extra flags. Run this once per machine.
Inputs
- Optional argument: an API key. If the user passed one with
/getlark:setup <key>, use it directly. Otherwise direct them to the dashboard to generate one.
Procedure
Step 1 — Verify prerequisites
Run in parallel:
node --version(must be ≥ 18)npm --versioncommand -v getlark || echo "not installed"
Report findings in one line. If Node is < 18 or missing, stop and tell the user to install Node 18+ first.
Step 2 — Install the CLI
If getlark is not on PATH:
npm install -g @getlark/cli
The package @getlark/cli installs a binary called getlark (not larkci).
If getlark is already installed, run getlark --version and ask if the user wants to upgrade (npm update -g @getlark/cli). Do not upgrade without confirmation.
Verify the install:
getlark --version
Step 3 — Obtain an API key
If the user did not provide a key:
Open https://dashboard.getlark.ai/settings/api-keys, create a new key, and paste it here.
Wait for the user to paste. Never echo the full key back — truncate to sk_...abcd when confirming.
Step 4 — Authenticate
Prefer getlark login — it saves credentials to ~/.getlark/config.json without touching shell rc files:
getlark login --api-key <KEY>
If the user prefers an environment variable (e.g. for CI or shell-wide access), export GETLARK_API_KEY in their rc file instead. Detect the shell:
echo "$SHELL"
Map to rc file:
/bin/zshor/usr/bin/zsh→~/.zshrc/bin/bash→~/.bashrcon Linux,~/.bash_profileon macOS/usr/bin/fish→~/.config/fish/config.fish
Check whether GETLARK_API_KEY is already exported in that file before appending. If not, append:
# For bash/zsh:
echo '\nexport GETLARK_API_KEY="<KEY>"' >> ~/.zshrc
# For fish:
echo '\nset -gx GETLARK_API_KEY "<KEY>"' >> ~/.config/fish/config.fish
Use the Edit or Write tool rather than echo when possible, so the diff is visible. Never commit or stage rc files.
Tell the user to either source the rc file or open a new shell. Do NOT run source inside the Bash tool — it will not affect the user's interactive shell.
Note: LARKCI_API_KEY is accepted but deprecated — the CLI will warn to rename it to GETLARK_API_KEY.
Default (https://api.getlark.ai) covers all production users.
Step 5 — Smoke test
Run:
getlark workflows list --limit 1
A JSON response with a workflows array (possibly empty) confirms setup. A 401/403 means the key is wrong; a network error means GETLARK_API_URL is wrong.
Success criteria
getlark --versionprints a version- Authentication works: either
getlark loginwas run orGETLARK_API_KEYis in the environment getlark workflows list --limit 1returns a JSON response
Recovery
- Command not found after install: check
npm bin -gis onPATH; suggestexport PATH="$(npm bin -g):$PATH"or reinstalling withsudoon Linux. - 401 Unauthorized: the pasted key is wrong; send the user back to https://dashboard.getlark.ai/settings/api-keys.
- ENOTFOUND / network error: confirm they're not behind a proxy that blocks
api.getlark.ai.