Spaces: A CLI Built for Humans and Agents
Mistral says Spaces became an agent-ready CLI by making every interactive choice available as flags, defaults, or structured context.
The post says Spaces began as an internal platform tool for scaffolding projects, running dev environments, generating configs, and deploying to staging. Mistral found that TUIs and hidden assumptions like current working directory blocked coding agents, so it added headless paths such as `--components`, `-y`, explicit config discovery, and JSON-readable plugin data. It also generates `context.json` and `AGENTS.md` so agents can read project state and operating rules before acting. The claim is that these changes made the CLI more composable and testable for humans as well as usable by agents. Mistral AI's note
The post says Spaces began as an internal platform tool for scaffolding projects, running dev environments, generating configs, and deploying to staging. Mistral found that TUIs and hidden assumptions like current working directory blocked coding agents, so it added headless paths such as `--components`, `-y`, explicit config discovery, and JSON-readable plugin data. It also generates `context.json` and `AGENTS.md` so agents can read project state and operating rules before acting. The claim is that these changes made the CLI more composable and testable for humans as well as usable by agents. Mistral AI's note
score 5