How to use the Paperclip Plan-to-Task Converter agent skill

What the skill makes the agent do

The skill tells the agent to decide whether delegation is justified before creating tasks. Small, cohesive work stays on its current ordinary task with one end-to-end owner. A plan's steps, files and phases do not each automatically become a task.

A separate task is created only for a useful boundary: a different owner or permission, independent parallel output, a real prerequisite, independent review, or substantial follow-up that needs its own lifecycle.

• Check the available agents before assigning work.
• Give each delegate enough context and a clear deliverable to work independently.
• Link dependent tasks with blockedByIssueIds (or the native dependency tool), then verify the saved relationships.
• Reviewers record findings on their own review task and complete it, including an adverse verdict.
• Honor existing approval boundaries. Decomposition does not create new approval gates.

Install the skill from the ZIP

Download the ZIP for paperclip-converting-plans-to-tasks from the page where you found this guide, then unzip it. The skill's only file is SKILL.md.

For a personal install, place the unzipped skill folder in ~/.claude/skills/. To share it with one project, place it in that project's .claude/skills/ folder.

Alternatively, upload the ZIP in the Claude apps under Settings → Capabilities → Skills.

• Personal: ~/.claude/skills/
• Project: .claude/skills/ inside your project
• Claude apps: Settings → Capabilities → Skills, then upload the ZIP

Trigger the skill with natural requests

You do not need a special command. The skill's description says to use it when turning an authorized plan into work, so ask for that in plain language. Mention the plan and the outcome you want, such as separate owners, dependencies or a review task.

Be explicit that the plan is already approved. The skill does not grant permission to start work.

• "Turn the approved migration plan into Paperclip tasks, keeping small steps together."
• "Split this authorized plan into tasks where a separate owner or real prerequisite requires it, and link the dependencies."
• "Create a review task for the security reviewer and make sure the findings go on that task."

What you need before you use it

The skill assumes a running Paperclip instance with its task, agent and dependency tools. The instructions depend on Paperclip's task, dependency and approval mechanics, so the skill does little on its own outside that environment.

• A running Paperclip instance
• Access to its task and agent tools
• The dependency field blockedByIssueIds, or the native dependency tool

Safety notes: scripts and network

According to the skill's listing, it runs no scripts and uses no network. It consists of one instruction file, SKILL.md. Its effects come from the Paperclip tools the agent calls, so review the tasks and dependencies the agent creates, especially when assigning work to other agents.

• Runs scripts: no
• Uses network: no
• Files: SKILL.md

Tips for better results

Give the agent the full authorized plan and say which parts are already approved. Ask it to justify each separate task against the boundaries in the skill. After it finishes, check the dependency links, since the skill requires verifying saved relationships. Use dependencies only for true prerequisites so independent work can start immediately.

• Provide the complete plan and its approval status.
• Ask why each separate task was created.
• Confirm dependencies were saved as intended.
• Keep dependency chains short to avoid needless waiting.