tel

Route whitelisted third-party service actions through the Trusted Execution Layer.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/Z5Jonathan-maker/ai-coding-os --skill tel
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tel
Source: https://github.com/Z5Jonathan-maker/ai-coding-os/tree/main/claude/skills/tel
Command: npx skills add https://github.com/Z5Jonathan-maker/ai-coding-os --skill tel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

TEL prevents credentials from entering Claude chat transcripts while still enabling authorized, audited actions against third-party services on the user’s behalf.

Core Features & Use Cases

  • Credential-safe execution: Executes whitelisted TEL actions using credentials stored in 1Password, ensuring Claude never sees the secret.
  • Strict allowlisting with auditability: Enforces policy YAML whitelisting, captures audit_id for traceability, and supports undo tokens for reversible actions.
  • Operational controls: Provides health checks, service/action listing, dry-run validation (request-shape checks without execution), and policy reload for updates.

Use when you need to create or manage resources in services like Gamma, GitHub, Gmail, Vercel, Notion, or Linear through credentialed operations without exposing OAuth tokens or API keys in chat.

Quick Start

Ask to use TEL to perform the credentialed action, then provide the exact service, action, and arguments you want executed through the TEL client.

Frequently Asked Questions about tel

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

FAQPage Schema
How do I execute credentialed API actions without exposing secrets in chat transcripts?

Credentialed API actions are routed through the Trusted Execution Layer (TEL), which executes whitelisted requests using secrets stored in 1Password. This ensures credentials never enter Claude chat transcripts while still performing authorized operations on your behalf.

What is the Trusted Execution Layer and how does it secure third-party service operations?

The Trusted Execution Layer (TEL) is a secure routing mechanism that enforces strict API allowlisting and auditability for third-party service operations. It captures an audit_id for traceability and supports undo tokens for reversible actions across services like GitHub, Gmail, and Notion.

Can I manage GitHub, Gmail, or Notion resources without passing OAuth tokens directly to the model?

Yes, you can manage resources in GitHub, Gmail, Notion, Vercel, and Linear without passing OAuth tokens to the model. TEL routes these credentialed actions through an execution layer that retrieves secrets from 1Password securely.

How do I validate a credentialed API request shape before actually executing the action?

You can validate a credentialed API request shape by performing a dry-run validation through TEL. This process checks the request structure against local policy YAML files without executing the action, ensuring the operation is allowed before committing.

Do I need to define a policy YAML file before running credentialed actions through TEL?

Yes, a local TEL policy YAML file is required before running credentialed actions. TEL enforces strict allowlisting by checking for policy existence and verifying TEL health before executing any whitelisted requests or capturing audit IDs.