"Discover and operate the LeptonAI lep CLI safely, including
Scanned 9/8/2026
Install to Claude Code
npx -y skills add VectorSpaceLab/AREX-Skill --skill cli-operations --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Cli Operations?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/vectorspacelab-cli-operations-ce4a5d00)More formats (shields.io, HTML) on the badges page.
---
name: cli-operations
description: "Discover and operate the LeptonAI lep CLI safely, including
command-tree help, aliases, abbreviations, output handling, and read-first
mutation gates."
disable-model-invocation: true
metadata:
disco-role: operating
license: Apache 2.0
---
# CLI Operations
Use this sub-skill when the task is about the global `lep` command line itself: discovering commands, checking the installed CLI version, orienting to command groups, interpreting help output, handling hidden aliases or click abbreviations, scoping logs/output, or planning whether a command is read-only, mutating, or destructive.
Do **not** use this sub-skill for credential setup details, resource-specific create/update recipes, or Python SDK calls. Route those to the neighboring workspace/auth, workload-management, storage/secrets/ingress, or SDK sub-skills after this sub-skill has established the safe CLI plan.
## Load These References
- Start with [references/cli-command-map.md](references/cli-command-map.md) to discover top-level groups, hidden compatibility aliases, and abbreviation behavior.
- Use [references/safety-and-output.md](references/safety-and-output.md) before any command that can mutate cloud resources, local CLI state, remote file storage, or local files.
- Use [references/troubleshooting.md](references/troubleshooting.md) when `lep` is missing, help is ambiguous, a blank option is rejected, a version warning appears, read-only commands hit auth/network errors, or destructive confirmation is unclear.
- For a safe local command-tree snapshot, run the bundled script:
```bash
python scripts/lep_help_snapshot.py --groups endpoint workspace job pod ingress storage file --json
```
The script runs only `lep --version`, `lep --help`, and selected `lep <group> --help` commands with a timeout; it strips common credential environment variables and does not call live workspace APIs.
## Operating Protocol
1. Verify the installed CLI surface with `lep --version`, `lep --help`, and the relevant `lep <group> --help` before relying on a command or flag.
2. Expand hidden aliases and abbreviations to full command names in any plan or final command. Do not use abbreviated commands for mutations.
3. Classify the requested operation: help/version, read-only, local-state mutation, cloud mutation, destructive cloud mutation, remote file mutation, local file write, or interactive session.
4. For reads, use the narrowest command that answers the question and avoid token-printing commands unless the user explicitly asks for them.
5. For every mutation or destructive operation, read current state first when possible, then present the exact command, target resource, intended workspace, observed state, and one-sentence impact. Execute only after explicit confirmation for that single target.
6. For logs and file-output options, scope by resource name, replica/job/time/query/path before running, and confirm local write destinations.
7. If help output or the installed version disagrees with these references, treat the installed `lep` help as authoritative and update the plan accordingly.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!