Skip to main content

Package recurring tasks as Skills / Slash Commands

Package recurring tasks as Skills / Slash Commands

Tags: Automation | Skills | Slash commands | Team workflows

1. Purpose​

Use this workflow when the same multi-step task is performed repeatedly. CodeVista can package the task as a reusable Skill or Slash Command that preserves the agreed sequence, inputs, permissions, safety rules, and output format.

2. Before You Begin​

  • Confirm that Skills or Slash Commands are supported in the current CodeVista version.
  • List the recurring steps and decide which actions may read, write, or run commands.
  • Choose the project location, expected inputs, and required output.

3. Starter Prompt​

Create a reusable [Skill or Slash Command] named [command-name].

Purpose:
[What the workflow accomplishes]

Workflow:
1. [Recurring step 1]
2. [Recurring step 2]
3. [Recurring step 3]

Inputs:
- [Required input 1]
- [Required input 2]

Permissions:
- Allowed: [read, write, or command actions]
- Prohibited: [actions that require approval or must never run]

Output:
[Required result and format]

Save it in the appropriate project location, explain how to invoke it, and validate it once if possible.

4. How to Use​

  1. Write the recurring workflow in the order it should run.
  2. Choose a short, memorable name and define its required inputs.
  3. State the allowed actions, approval boundaries, and prohibited actions.
  4. Ask CodeVista to package and save the workflow.
  5. Review the generated definition and run one representative validation.
  6. Commit or share the approved workflow with the team when appropriate.

5. Expected Result​

  • Reusable trigger: A short command starts an established workflow without rewriting the full prompt.
  • Consistent execution: The same steps, constraints, and output structure are applied each time.
  • Team reuse: A project level workflow can be reviewed, versioned, and shared with contributors.

Package recurring tasks as skills and slash commands demonstration

6. Tips​

  • Keep each Skill focused on one repeatable outcome.
  • Make permissions explicit and never embed credentials or secrets in the definition.
  • Include one realistic example so the expected behavior is easy to validate.