Guide · Agent skill
How to use the Create Runner Adapter agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The create-runner-adapter skill is a set of written instructions that an AI agent follows when you ask it to add a new AI provider or coding-tool integration, called an adapter, to Paperclip's native Runner system. It is written for maintainers of the paperclipai/paperclip repository and developers who need a structured acceptance checklist. After this guide you will know what the skill does, how to install it, how to phrase a request that triggers it, and what it needs before it can be useful.
Create Runner Adapterpaperclipai/paperclip · MIT Download the skill Quick start
- 1Get the Paperclip checkout. Clone or open the paperclipai/paperclip repository and confirm it contains packages/paperclip-runner and tests/runner-e2e.
- 2Install the skill. Unzip the downloaded ZIP into ~/.claude/skills/ for personal use or into .claude/skills/ in the project, or upload it under Settings → Capabilities → Skills in the Claude apps.
- 3Gather provider access. Make sure you have access to the target provider's harness and account for real end-to-end testing, and decide your spending limits.
- 4Make the request. Ask the agent in plain language to add a native Runner adapter for a named provider or coding tool.
- 5Let the agent define the release. Review the recorded user outcome, platforms, setup, capabilities, acceptance model and delivery target before investigation begins.
- 6Review evidence at each stage. Check the investigation notes, the single end-to-end slice results and the acceptance ledger rather than trusting a success message.
- 7Hand off using the templates. Confirm the handoff covers repository checks, passing evidence, remaining limits, installation, operation and rollback.
What the create-runner-adapter skill makes the agent do
An agent skill is a folder with a SKILL.md file that tells an agent how to handle a particular kind of task. This one covers adding a new harness or provider integration to Paperclip's new, native Runner execution system. Its instructions cover the full path: defining a finite release, investigating the target harness, proving one complete end-to-end slice with real tasks, handling events, approvals, recovery and cancellation, verifying packaged delivery, and handing off documented results.
The skill does not contain the detailed rules itself. It points to the repository's Runner integration playbook and working templates for the release contract, capability inventory, acceptance ledger and failure record. Legacy direct adapters are explicitly out of scope, so the skill will not be used to edit or migrate existing adapters.
Install the skill with a ZIP download or the Claude apps
Download the ZIP from the skill's page on this directory. To install it for your user account, unzip it into ~/.claude/skills/. To install it for a single project, unzip it into that project's .claude/skills/ folder.
If you use the Claude apps instead, upload the ZIP under Settings → Capabilities → Skills.
- Personal install: ~/.claude/skills/
- Project install: .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills
Check your checkout before you start
The skill depends on files in the paperclipai/paperclip repository. Before the agent edits anything, it should locate the target checkout from the current workspace or git rev-parse --show-toplevel, then confirm that packages/paperclip-runner and tests/runner-e2e exist there. It should also follow that checkout's AGENTS.md and scoped rules.
If you installed the skill outside the Paperclip codebase, it cannot work on its own, because the playbook and templates it relies on live in that repository.
git rev-parse --show-toplevelTrigger the skill with a natural request
You do not need a special command. Describe the work you want done in plain language that mentions adding a new Runner adapter or harness. The skill's description covers adding a harness or provider integration to native Paperclip Runner, from requirements and protocol research through implementation, packaging and real end-to-end acceptance.
Name the provider or tool and the outcome you need, so the agent knows which harness to investigate. If you are asking about a legacy direct adapter, expect the skill not to apply.
- Add a native Runner adapter for [provider] to the Paperclip Runner system.
- Start the create-runner-adapter workflow for a new [coding tool] harness.
- Plan the acceptance testing for a new Runner harness integration.
Follow the seven-stage work sequence
The skill directs the agent through seven stages in order. The agent first defines a finite release, recording the user outcome, supported platforms, setup and authentication, required capabilities and delivery target. Next it investigates the real harness using official docs, source and protocol traffic, then proves one complete slice with a real Paperclip task whose output is checked independently.
After that come the remaining behaviour (events, approvals, recovery, cancellation), delivery proof using clean installed packages, release qualification, and a handoff covering installation, operation and rollback. The skill keeps merge, publish, deploy and canary claims separate.
Understand the safety notes
This skill does not run scripts and does not use the network by itself. The work it describes does involve real provider accounts for end-to-end acceptance testing, so you need access to the target provider's harness and account. The skill warns that spending limits should be recorded during release definition.
The skill also states that a green unit suite, a screenshot or an agent's own success claim does not count as proof. Expect the agent to check files, artifacts, state and cleanup independently.
What people use it for
- Adding a new coding-tool harness to the native Paperclip Runner system as a repository maintainer.
- Building a structured acceptance checklist for a new Runner provider integration.
- Preparing a release contract, capability inventory and acceptance ledger from the repository templates.
- Checking that a new integration's installed package works in the agreed execution environment before release.
Frequently asked questions
What is the create-runner-adapter skill?
It is an agent skill, a SKILL.md instruction file, that guides an AI agent through adding a new provider or coding-tool integration to Paperclip's native Runner system. It covers requirements, investigation, end-to-end testing, delivery and handoff. It points to repository documents for the detailed rules.
Do I need to pay to use this skill?
The skill is published under the MIT licence. The source does not state prices for the provider accounts you would need for real end-to-end testing, so check those separately with each provider.
What do I need before installing it?
You need a checkout of the paperclipai/paperclip repository that contains packages/paperclip-runner and tests/runner-e2e, plus access to the target provider's harness and account for real acceptance testing.
Where do I put the skill files?
Unzip the ZIP into ~/.claude/skills/ for your user account or into .claude/skills/ in a project. In the Claude apps, upload the ZIP under Settings → Capabilities → Skills.
Can I use this skill to migrate an old adapter?
No. The skill explicitly excludes editing or migrating legacy direct adapters, because that work could disturb existing agents or records. It is only for adding new native Runner adapters.
Are there alternatives or does it compare to other tools?
The source does not compare this skill with other tools or list alternatives, so no comparison is claimed here. Its scope is narrow: adding native Runner adapters inside the Paperclip repository.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you maintain the Paperclip repository or are adding a new native Runner harness and want a disciplined, checklist-driven process with real end-to-end testing. Pick something else if you are working outside the Paperclip codebase, need to migrate a legacy direct adapter, or do not have access to the provider account needed for acceptance testing.
Download the skill