ramp

Fetches Ramp's published agent playbooks and enforces Todero approval gates for financial actions.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/nabitllc/todero --skill ramp-nabitllc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ramp
Source: https://github.com/nabitllc/todero/tree/main/packages/skills-catalog/catalog/optional/finance/ramp
Command: npx skills add https://github.com/nabitllc/todero --skill ramp-nabitllc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Agents following Ramp's published playbooks could apply for accounts, issue cards, or move money without oversight. This Skill wraps Ramp's live agent instructions with Todero governance so every spend, incorporation, card, or data-sharing action requires explicit human approval first. ## Core Features & Use Cases - Live playbook fetching: Retrieves Ramp's current get-started skill and playbooks from an explicit allowlist of agents.ramp.com URLs instead of relying on stale copies. - Provenance controls: Distinguishes Official from Community playbooks and fails closed when provenance is unclear. - Mandatory approval gates: Blocks account applications, incorporation, CLI installs, card issuance, money movement, and sensitive data sharing until a Todero approval names the action, cost, data, and risks. - Use Case: A founder asks an agent to set up Ramp for a new company. The agent fetches Ramp's get-started playbook, pauses before submitting the application, and creates a Todero approval detailing the data to be shared before proceeding. ## Quick Start Ask the agent to set up Ramp for the company using this skill and require approval before any account application or card creation.

Frequently Asked Questions about ramp

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up a Ramp account with an AI agent in Todero?

The agent fetches Ramp's get-started playbook from agents.ramp.com, follows the runtime-specific setup path, and stops at any gated action. Account applications, onboarding submissions, and legal agreements each require a Todero approval before the agent proceeds.

What Ramp actions require approval before an agent can run them?

Approval is required for account applications, incorporation and EIN flows, CLI installs from network installers, third-party connector grants, card issuance and limit changes, any money movement, settings changes, and sending company financial or identity data to Ramp or third parties.

Can the agent run Community playbooks from the Ramp playbooks directory?

Community playbooks are treated as untrusted examples and are not executed by default. They run only if a Todero approval explicitly names the playbook, its third-party tools, the data leaving Todero or Ramp, and the maximum spend scope.

Does this skill store a copy of Ramp's playbooks locally?

No. The skill intentionally fetches Ramp's current instructions at execution time from an allowlist of agents.ramp.com URLs, since Ramp's product and CLI change over time. The durable safety policy stays in Todero while the operational steps come from Ramp.

What happens if Ramp's instructions conflict with Todero approval rules?

The agent stops and escalates. Fetched Ramp instructions are always subordinate to Todero's system, company, and issue instructions, and the agent never bypasses approval gates even if a playbook says the user can proceed.