From getlark
This skill should be used when the user asks to "list workflows", "show getlark workflows", "get workflow details", "archive a workflow", "update a workflow", "list workflow groups", "manage secret contexts", "show executions", "show repairs", "show generations", "show events", or runs `/getlark:manage`. Covers read/update/archive operations across all getlark resources and formats CLI JSON as human-friendly tables. Use `create-workflow` to create new workflows and `invoke-workflow` (or `validate-branch`) to run them — this skill never triggers executions, it only inspects and mutates metadata.
How this skill is triggered — by the user, by Claude, or both
Slash command
/getlark:manage [resource] [action] [args...][resource] [action] [args...]This skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Inspect and mutate getlark resources via the `getlark` CLI, then render JSON output as a readable table for the user. Applies to: workflows, workflow-groups, secret-contexts, executions, repairs, generations, events, jobs, and config profiles.
Inspect and mutate getlark resources via the getlark CLI, then render JSON output as a readable table for the user. Applies to: workflows, workflow-groups, secret-contexts, executions, repairs, generations, events, jobs, and config profiles.
| Resource | Actions |
|---|---|
| workflows | list, get, update, archive, unarchive |
| workflows/executions | list, get, logs, cancel |
| workflows/repairs | list, get, trigger, cancel |
| workflows/generations | list, get, logs, cancel |
| workflows/events | list |
| workflow-groups | list, get, create, update, delete |
| secret-contexts | list, get, create, update, delete, delete-key |
| jobs | list, get, cancel |
| config (profiles) | list, use |
Creation of workflows lives in /getlark:create-workflow, not here. Invocation lives in /getlark:invoke-workflow. Bulk workflow imports (jobs upload/validate) live in /getlark:manage-bulk.
Examples:
| User request | Command |
|---|---|
| "list my workflows" | getlark workflows list --limit 100 |
| "show workflow wf_abc" | getlark workflows get wf_abc |
| "archive that workflow" | getlark workflows archive <id> (confirm first) |
| "rename workflow X to Y" | getlark workflows update <id> --name "Y" |
| "list workflow groups" | getlark workflow-groups list |
| "what secrets are in the staging context" | getlark secret-contexts get staging |
| "show last execution of wf_abc" | getlark workflows executions list wf_abc --limit 1 → executions get |
| "show repair history for wf_abc" | getlark workflows repairs list wf_abc |
| "list jobs" | getlark jobs list |
| "list pending jobs" | getlark jobs list --status pending |
| "check status of job job_abc" | getlark jobs get job_abc |
| "cancel job job_abc" | getlark jobs cancel job_abc (confirm first) |
| "list my profiles" | getlark config list |
| "switch to staging profile" | getlark config use staging |
If the user refers to a resource by name (not ID), resolve via the corresponding list call first.
For archive, unarchive, update, delete, delete-key, cancel, trigger: echo the exact command and the resource identifier and ask the user to confirm before running. Deletes and archives are user-visible and not always reversible.
list and get run without confirmation.
The CLI emits pretty-printed JSON. Convert it to a compact, human-readable table before replying to the user. Guidance per resource:
workflows list — columns: id, name, status, mode, group_id, last_execution_result_type, updated_at
workflows get — key/value list, highlight: status, mode, secret_contexts, group_id, schedule, last_execution_*, last_repair_*.
workflow-groups list — id, name, updated_at
secret-contexts list — context, updated_at
secret-contexts get — context + bullet list of keys
executions list — id, status, started_at, stopped_at (duration if both present)
executions get — highlight status, summary; list steps[] with step, result, and any friction_points
repairs list / generations list / events list — id, status, started_at, stopped_at
jobs list — id, name, type, status, created_at; include dashboard_url as a clickable link
jobs get — key/value, highlight status, type; always include dashboard_url as a link
config list — table of profiles: mark current with *, columns name, api_url (or "(default)"), masked api_key
Drop timestamps that are null and ISO-trim to minute precision when space is tight. For long lists (>15 rows), show top N and note total.
Include the relevant dashboard URL after the table so the user can click through:
https://dashboard.getlark.ai/workflows/<id>https://dashboard.getlark.ai/jobs/<id> (or use the dashboard_url field from the API response directly)list subcommands accept --limit (max 100) and --offset. If a response has has_more: true, offer to fetch the next page — do not auto-loop.
list first.npx claudepluginhub getlark/skills --plugin getlarkProvides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.