The Syntic Code CLI
The CLI is the syntic command. Run it in a project folder and you get an interactive screen where you chat with the agent, approve its actions and watch it work. Add -p and it runs one task without the screen, which is how you use it in scripts and CI.
syntic # interactive session in this folder
syntic "why does npm test fail?" # start with a first message
syntic -p "summarize src/router.ts" # one task, print the answer, exit
syntic --continue # pick up the latest session hereIn this section
| Page | Covers |
|---|---|
| Install & update | Installers for macOS, Linux and Windows, Homebrew, syntic update, removal. |
| Sign in | Browser sign-in, signing in over SSH or in CI, checking and clearing credentials. |
| The terminal interface | What is on screen, attaching files, inline versus full-screen, accessibility. |
| Keyboard shortcuts | Every key, vim mode, and how to rebind keys. |
Shell mode (!) | Running your own commands without leaving the session. |
| Sessions & resume | Continuing, naming, branching and cleaning up conversations. |
| Context & compaction | What fills the context window and how to keep room. |
| Automatic checks | The Checked: line after bigger changes, and how sub-agent reports are checked. |
| Everyday tasks | Built-in commands for reviews, commits and pull requests, plus prompt ideas. |
| Prompt library | Ready-to-copy prompts grouped by task. |
| Best practices | Habits that get better results: verification, planning, context, scaling up. |
| Memory (SYNTIC.md) | Instruction files the agent reads every session. |
Shared topics
Permissions, memory, subagents, skills, hooks, MCP servers and rewind work the same in every Syntic Code surface, so they have their own topics: Permissions & safety, Agents & automation, Extend and Configure. Every flag and subcommand is listed in the CLI reference.