Commet cli
Operate the installed Commet CLI for project setup, local diagnostics, agent rules, organization linking, configuration sync, webhook forwarding, scaffolding, and resource commands. Use the CLI's installed capabilities and version-matched documentation instead of a copied command inventory, and require explicit organization and sandbox or live context before remote effects.From its SKILL.md
npx -y skills add commet-labs/skills --skill commet-cliAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 3 stars3 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.
SKILL.md
3.1 KB, as published. Nobody here has run it
Commet CLI
Use the installed CLI as the command contract. Do not rely on a command or flag list embedded in this skill.
Inspect first
- Resolve the executable the project actually uses and record its version.
- Run
commet --output agentfor the installed capability manifest andcommet <command> --helpfor the exact command being considered. - When
@commet/nodeor another Node integration is installed, readnode_modules/@commet/node/docs/manifest.json, its entrypoint, and the installed CLI documentation it indexes. - Run
commet doctor --output agentfrom the project root and address failed package, documentation, compatibility, API-version, or project-context checks before changing the integration.
If the installed CLI does not expose a command or flag, do not substitute one from this skill, training data, or the latest website.
Choose the workflow
- For local diagnosis, run doctor and report its structured evidence without printing secret values.
- For agent instructions, use the installed
agentscapability. Setup mutates repository files, so run it only when the user explicitly asks. - For configuration sync, inspect the installed pull or push capability, preview the exact diff when supported, and preserve project-owned configuration.
- For webhook forwarding, confirm the local destination, requested event scope, and process lifetime before starting it.
- For scaffolding, confirm the destination directory and organization context before creating files or remote resources.
- For resource operations, derive resources, actions, parameters, and output format from the installed capability manifest and help output.
Guard effects
Before a command that writes local files, state the exact target and require the user's request to authorize that local mutation.
Before a command that links a project, pushes configuration, creates or mutates a remote resource, or forwards remote events:
- State whether the effect is remote or both local and remote.
- Require the exact organization and
sandboxorlivemode. Use a passingPROJECT_CONTEXT_VALIDdoctor check when available. - Show the documented preview or dry run when the installed command supports one.
- Do not perform a live write unless the user's request explicitly authorizes it.
Authentication commands may open a browser or store credentials. Explain that effect before running them. Never print credentials or secret values, and never invent a non-interactive flag that the installed CLI does not advertise.
Verify
After the requested operation, use structured output or a documented read command to verify the result. For repository writes, show the resulting diff and preserve unrelated user changes.