Skip to content
Current section: Get Started

Start Fresh or Remove cedr

Two prompts in the Ask your agent menu on the app’s Overview cover a whole machine: Start fresh and Remove cedr from this machine. Both list everything first, back up configuration before deleting it, wait for your approval on each item and report what is left. To remove one product only, use the Remove section on its setup page: Route, Council, Guard, Sync or Monitor.

Neither prompt touches your organization’s data in the app: its members, repos, picks, usage and saved settings stay. Revoke keys on the API keys page, connected applications on the MCP page, and delete an organization or your account in the app (Delete Your Account).

  • Settings in ~/.config/cedr/: routes.toml, monitor.toml, shared routes in org-routes/ and the api-key file.
  • State in ~/.local/state/cedr/: route history, plan readings, the Monitor store and machine-id.
  • A cache in ~/.cache/cedr/.
  • Integrations: the Monitor timer, Claude Code’s status line, the ## Cedr Route section in ~/.claude/CLAUDE.md and ~/.codex/AGENTS.md, and the cedr and cedr-cloud MCP servers.
  • In each repository: .cedr/ and .github/workflows/cedr.yml.

The paths on this page are the defaults. If XDG_CONFIG_HOME, XDG_STATE_HOME or XDG_CACHE_HOME is set, cedr keeps these folders under $XDG_CONFIG_HOME/cedr, $XDG_STATE_HOME/cedr and $XDG_CACHE_HOME/cedr instead, and the prompts below check for that first.

Start fresh clears the machine’s cedr setup but keeps the CLI, then sets cedr up again. It asks before deleting ~/.config/cedr/api-key (keeping it saves creating a new key) and ~/.local/state/cedr/machine-id (deleting it makes the machine show up as a new one in the app).

Help me start fresh with cedr on this machine: clear its local setup, keep the CLI, then set it up again.
Before deleting anything, list every file, setting, registration and scheduled job you would remove, with its path, and wait for my approval on each one. Skip anything I want to keep. Back up every configuration file you will change or delete into a new folder ~/cedr-backup-<YYYY-MM-DD>/ first, keeping its path. Never touch my organization's data in the cedr.dev app. Finish by telling me what you removed, where the backups are and what is left.
1. Back up ~/.config/cedr/ (routes.toml, monitor.toml, org-routes/ and the api-key file), ~/.local/state/cedr/ (route history, readings, the Monitor store, machine-id) and ~/.cache/cedr/ into the backup folder.
2. Undo these integrations, one approval each:
- the Monitor timer: remove it with cedr monitor schedule --remove;
- Claude Code's status line in ~/.claude/settings.json: if it is exactly cedr route statusline, remove the statusLine setting; if it wraps another command (cedr route statusline -- <command>), put back just that command;
- the ## Cedr Route section in ~/.claude/CLAUDE.md and ~/.codex/AGENTS.md;
- the MCP servers named cedr and cedr-cloud in each agent (for example claude mcp remove cedr, codex mcp remove cedr, or deleting the entry from the agent's MCP config).
3. Clear the settings and state folders above. Ask me before deleting ~/.config/cedr/api-key (keeping it saves creating a new key) and ~/.local/state/cedr/machine-id (deleting it makes this machine show up as a new one in the app).
4. Run cedr --version to confirm the CLI still works, then set cedr up again following https://cedr.dev/docs/agents.

This removes every product’s local setup, then the CLI itself, the way it was installed. Repositories keep their own cedr files until you clean each one up.

Remove cedr from this machine completely.
Before deleting anything, list every file, setting, registration and scheduled job you would remove, with its path, and wait for my approval on each one. Skip anything I want to keep. Back up every configuration file you will change or delete into a new folder ~/cedr-backup-<YYYY-MM-DD>/ first, keeping its path. Never touch my organization's data in the cedr.dev app. Finish by telling me what you removed, where the backups are and what is left.
1. List and back up ~/.config/cedr/ (routes.toml, monitor.toml, org-routes/ and the api-key file), ~/.local/state/cedr/ (route history, readings, the Monitor store, machine-id) and ~/.cache/cedr/.
2. Undo these integrations, one approval each:
- the Monitor timer: remove it with cedr monitor schedule --remove;
- Claude Code's status line in ~/.claude/settings.json: if it is exactly cedr route statusline, remove the statusLine setting; if it wraps another command (cedr route statusline -- <command>), put back just that command;
- the ## Cedr Route section in ~/.claude/CLAUDE.md and ~/.codex/AGENTS.md;
- the MCP servers named cedr and cedr-cloud in each agent (for example claude mcp remove cedr, codex mcp remove cedr, or deleting the entry from the agent's MCP config).
3. Run cedr telemetry off, which stops usage data and deletes the install ID and queued events.
4. Delete the settings, state and cache folders above.
5. Repositories keep their own cedr files (.cedr/, .github/workflows/cedr.yml). Ask me which repos to clean up; for each, follow the Remove Guard and Remove Sync steps in the docs (https://cedr.dev/docs) on a new branch.
6. Uninstall the CLI the way it was installed: run command -v cedr and show me the path. If it is ~/.cargo/bin/cedr, run cargo uninstall cedr; otherwise delete that file (on Windows, cedr.exe in the folder it shows). Then confirm cedr no longer runs.
  1. Run cedr monitor schedule --remove and cedr telemetry off.
  2. In ~/.claude/settings.json, remove a statusLine of exactly cedr route statusline, or put back the command it wraps.
  3. Remove the ## Cedr Route section from ~/.claude/CLAUDE.md and ~/.codex/AGENTS.md, and the cedr and cedr-cloud MCP servers with your agent’s own command (for example claude mcp remove cedr-cloud).
  4. Back up, then delete ~/.config/cedr/, ~/.local/state/cedr/ and ~/.cache/cedr/.
  5. Find the CLI with command -v cedr. Run cargo uninstall cedr if it is ~/.cargo/bin/cedr; otherwise delete that file.