Who it’s for
- Teams that use coding agents such as Claude Code and Codex, and want every developer’s agent to follow the same company instructions, kept current without anyone copying files around.
- Teams that ship LLM features and want prompts out of application code, so a wording change is a publish rather than a deploy, with a history of every version.
- Engineering leads who want to read what their team’s Claude Code sessions did: the prompts, the tool calls, the tokens, and what that usage would cost at API prices.
What you get
Skills
A skill is a folder of instructions, with aSKILL.md entry document, that a coding agent reads from its skills directory. You store a skill in a project with the to11 CLI, and every machine that follows the project receives the new version on its next sync, in the directories Claude Code, Codex, and other agents read. See Skills.
Chat prompts
A chat prompt is a versioned template of messages, variables, tools, and model settings. You author it in the dashboard and your application renders its newest version at runtime with the TypeScript or Python SDK, then sends it with the provider client you already use. See Prompt management.Sessions
The to11 CLI uploads the transcripts of the Claude Code sessions you run, with secrets masked, to a project. The dashboard lists them and shows each one as a thread of prompts, responses, and tool calls, with its token usage and an API-price estimate. See Sessions.Access and accountability
Members sign in with GitHub, Google, or Microsoft and get access through roles at the organization, workspace, and project level. Applications and CI use scoped API keys. Changes are recorded in audit logs. See Platform.How it’s organized
Everything nests into Organization → Workspace → Project. Prompts, skills, environments, and sessions belong to a project. See Core concepts.Next steps
Quickstart
Create a project and use it from the SDK and the CLI.
Core concepts
How to11 is organized.