Todo in Agent

todo-in-agent is a Codex skill for turning useful but deferred agent discussions into prompt-backed todo items.

Use it when a discussion reaches a decision like "this is real, but not the priority right now" and you want to preserve the reasoning as a reusable follow-up prompt.

What It Produces

• A backlog title and priority
• The decision context
• Product or technical semantics that must be preserved
• Risks if the work is ignored
• A self-contained follow-up prompt for a future agent
• Acceptance criteria for the later pass

Install

Copy the skill folder into your Codex skills directory:

cp -R todo-in-agent "${CODEX_HOME:-$HOME/.codex}/skills/"

Then ask Codex to use todo-in-agent when you want to convert a discussion into a prompt-backed todo.

Example Request

This is lower priority now. Turn the discussion into a todo with a prompt so I can revisit it later.

Repository Layout

todo-in-agent/
  SKILL.md
  agents/openai.yaml

Open Source Compliance

License: MIT. See LICENSE.

Third-party dependencies: none.

Bundled assets: none.

Runtime network access: none. This repository contains a text-only Codex skill and metadata.

Security

Do not include secrets, credentials, private customer data, or proprietary source excerpts in public prompt-backed todos.

For reporting guidance, see SECURITY.md.

Contributing

See CONTRIBUTING.md.